Skip to main content

Full text of "Biology of Africanized and European honey bees, Apis mellifera, in Venezuela"

See other formats


BIOLOGY OF AFRICANIZED AND EUROPEAN HONEY BEES, 
Ap-ls men if era , IN VENEZUELA 



By 

ALAN B. BOLTEN 



A DISSERTATION PRESENfTED TO THE GRADUATE SCHOOL 

OF THE UNIVERSITY OF FLORIDA IN 

PARTIAL FULFILLMEm" OF THE REQUIREMEhrrS 

FOR THE DEGREE OF DOCTOR OF PHILOSOPHY 



UNIVERSITY OF FLORIDA 
1986 



"He must be a dull man who can examine the 

exquisite structure of a comb, so beautifully 

adapted to its end, without enthusiastic admiration." 

Charles Darwin 1859 



ACKNOWLEDGEMENTTS 

I would like to thank the members of my committee: Dr. Thomas C. 
Emmel, my chairman, for his continued support, guidance and 
encouragement throughout my graduate education; Dr. Malcolm T. Sanford 
for stimulating discussions and his thorough editing; and Dr. Jonathan 
Reiskind for his enthusiasm and helpful suggestions. I appreciate the 
comments made by Drs. James Nation and Frank Nordlie on the 
dissertation. I am also grateful to Professor Frank Robinson for 
introducing me to the excitement and challenges of honey bee research 
and management. Drs. John Harbo, Anita Collins and Tom Rinderer were 
excellent field companions, sharing their knowledge of honey bee 
research techniques, and creating a stimulating research environment, 
both in Venezuela and during my work in Baton Rouge. I particularly 
want to thank Dr. John Harbo for the instrumental inseminations and 
acknowledge his collaboration on both the bee size and egg laying rate 
experiments. I would also like to thank Dr. Orley Taylor for giving me 
the opportunity to study Africanized honey bees. 

This research was supported by the U.S. Department of Agriculture 
Cooperative Agreement No. 58-7B30-8-7 with the University of Kansas (0. 
R. Taylor, principal investigator). The Ministerio de Agricultura y 
Cria de Venezuela provided research facilities near Maturin. I would 
like to thank Med. Vet. Ricardo Gomez Rodriguez for his hospitality and 



ill 



logisti'c support. Laboratory facilities at the Universidad de Oriente 
in Jusepin were made available by Professor Dick Pulido. 

The research presented in this dissertation and the commitment to 
complete the writing could not have been accomplished without the 
collaboration, companionship, encouragement and insights of my wife, 
Karen Bjorndal, who shared not only the excitement and successes but 
also the frustrations and discomforts of Africanized honey bee research. 

Finally, I would like to thank my parents, who have always 
supported and encouraged my work. 



tv 



TABLE OF CONTENn'S 

Page 

ACKNOWLEDGEMENTS lii 

ABSTRACT vii 

CHAPTERS 

I INTRODUCTION 1 ' 

Evolutionary Origin and Distribution of Honey Bees 1 [ 

Importation of African Honey Bees Into Brazil and [ 

Their Dispersal Throughout South and Central America . . 1 } 

Characteristics of Africanized Honey Bees 3 j 

Purpose of My Research 6 j 

Identification of Honey Bees Used In My Research 7 j 

When and Where Research Was Conducted 7 

II WORKER BEE DEVELOPMENT TIMES 9 

Introduction 9 

Methods 14 

Results 18 

Discussion 20 

III irrrERACTION of maternal genotype, egg GENOTYPE AND COMB 

CELL SIZE ON HONEY BEE WORKER SIZE AND SIZE VARIATION ... 35 

Introduction 35 

Methods 38 

Results 41 

Discussion 42 

IV QUEEN DEVELOPMEhfT AND MATURATION 55 

Introduction 55 

Methods 57 

Results 63 

Discussion 65 



V QUEEN PUPAL WEIGHTS 79 

Introduction 79 

Methods 80 

Results 83 

Discussion 84 

VI EGG LAYING AND BROOD PRODUCTION RATES 

DURING THE FIRST BROOD CYCLE 92 

Introduction 92 

Methods 96 

Results 101 

Discussion 102 

VII SUCCESSFUL HYBRIDIZATION BETWEEN AFRICANIZED 
AND EUROPEAN HONEY BEES IN VENEZUELA WITH 

IMPLICATIONS FOR NORTH AMERICA 118 

Introduction 118 

Methods 123 

Results • 124 

Discussion 125 

VIII DISCUSSION: FACTORS CONTRIBUTING TO THE SELECTION 

ADVANTAGE OF AFRICANIZED HONEY BEES IN SOUTH AT^ERICA— 

THE RESOURCE UTILIZATION EFFICIENCY HYPOTHESIS 133 

Success of Introduced Populations of Honey Bees 133 

Factors Affecting Honey Bee Reproductive Rates 135 

Factors Contributing to the Selective Advantage 

of Africanized Honey Bees in South America 140 

Potential Impact of Africanized Honey Bees 

in North America 153 

APPE^vDICES 

A WORKER BEE DEVELOPMEm" TIMES AND MORTALITY 

DURING DEVELOPMENT 156 

B HONEY BEE SIZE, COMB CELL SIZE AND 

SIZE VARIATION . 159 

C CHANGES IN QUEEN PUPAL WEIGHT WITH AGE 165 

D ACCURACY OF TECHNIQUE USED TO ESTIMATE 

NUMBER OF BEES IN A COLONY 167 

LITERATURE CITED 168 

BIOGRAPHICAL SKETCH .... 182 



v1 



Abstract of Dissertation Presented to the Graduate School 

of the University of Florida in Partial Fulfillment of the 

Requirements for the Degree of Doctor of Philosophy 



BIOLOGY OF AFRICANIZED AND EUROPEAN HONEY BEES, 
Apis mellifera . IN VENEZUELA 

By 

Alan B. Bolten 

August 1986 

Chairman: Thomas C. Emmel 
Major Department: Zoology 

To determine factors responsible for the greater success of 
Africanized honey bees, Apis mellifera , in tropical regions of South 
America, demographic parameters affecting colony reproductive rates were 
evaluated for Africanized and European honey bees under identical 
conditions in Venezuela. Worker bee development time was evaluated as 
an interaction between egg genotype, comb cell size and nurse bee 
genotype. Africanized worker bees developed faster than European bees: 
18.9 and 19.8 days, respectively. There was no significant effect of 
comb cell size or nurse bee genotype. Mortality for different 
developmental stages was recorded. The relationship of worker bee 
development time to colony growth rate is discussed. 

Africanized queens develop in 14.5 days post-oviposit ion compared 
with 15.0 days for European queens. Queen pupal weights were not 
significantly different. Post-emergence maturation rates for 
Africanized and European queens were similiar as determined by both the 

vli 



ages when queens attracted drones and the ages when oviposit ton was 
initiated. 

Daily egg laying rates and brood production during initial colony 
growth were not significantly different for Africanized and European 
queens. Africanized and European worker bees did not differentially 
affect egg laying and brood production rates. 

Differences in reproductive rates between Africanized and European 
honey bees in South America cannot be attributed to differences in 
intrinsic demographic factors. A hypothesis based on differences in 
resource utilization efficiency is presented to explain the success of 
Africanized bees compared with European bees in South America. 

Results from reciprocal F-^ crosses indicate that bee size is a 
function of egg genotype, comb cell size and maternal genotype. The 
importance of maternal inheritance for reducing worker bee size 
variation within a colony is discussed. Advantages of smaller worker 
bee size are evaluated for Africanized bees. 

There are no effective reproductive isolating mechanisms operating 
between Africanized and European honey bee populations. Both 
Africanized and European queens mated with equal success with 
Africanized drones as measured by the numbers of spermatozoa in the 
spermatheca. The potential impact of Africanized bees on North America 
is analyzed with respect to hybridization and genetic introgression, 
resource competition, and selection advantages for European bees in 
temperate regions. 



vm 



CHAPTER I 
INTRODUCTION 



Evolutionary Origin and Distribution of Honey Bees 
Honey bees of the genus Apis have their greatest diversity in Asia 
(Michener 1979). Earliest fossil evidence for the genus is from 
Oligocene deposits in Europe (Zeuner and Manning 1976). The 
evolutionary relationships of the four generally recognized species of 
Apis are reviewed by Michener (1974). Three of the species (A. cerana . 
A. dors at a and A. florea ) are native only to Asia (Michener 1979; 
Ruttner 1975). The western honey bee (A. mell ifera ) is native to 
Africa* western Asia and Europe and may have evolved in tropical or 
subtropical Africa (Wilson 1971) or the Near East (Ruttner 1975). The 
widely different climatic conditions and floral resources under which 
populations of A. mell ifera evolved have resulted in a number of 
geographically recognizable subspecies (Alpatov 1929, 1933; Br. Adam 
1966; Dupraw 1965; Ruttner 1968, 1975, 1976a, 1976b; Smith 1961; Wafa, 
Rashad and Mazeed 1965). 

Importation of African Honev Bees into Brazil and Their Dispersal 
Throughout South and Central America 

European honey bees (A. mell ifera mell ifera and A. \n. li gustica ) 

had been introduced into Brazil by 1845 (Gerstaker cited in Pellet 1938; 

Woyke 1969). A. m. mell ifera is native to Europe in the regions west 

and north of the Alps and extending east into Central Russia; A. rn.. 

1 



2 

1 igustica is native to the Italian peninsula (Ruttner 1975). Because 
these European honey bee populations were not very successful in 
tropical and subtropical habitats of Brazil (Michener 1972), researchers 
believed that they could improve Brazil's honey production by breeding a 
honey bee better adapted to local conditions (Woyke 1969). With this 
intention, honey bee queens from South Africa (A. m. scutellata , 
formerly classified as adansonii , see Ruttner 1976a, 1976b, 1981) were 
imported into southeastern Brazil in 1956 (Kerr 1967). The following 
year, swarms escaped and hybridized with established European honey 
bees. The descendents from this hybridization are known as Africanized 
honey bees (Goncalves 1982). Details of the introduction and subsequent 
spread throughout South America have been extensively reviewed 
(Goncalves 1974, 1975, 1982; Kerr 1967; Michener 1972, 1975; Taylor 
1977, 1985; Taylor and Williamson 1975; Woyke 1969). 

In the 30 years since African honey bees were imported into 
southeastern Brazil, their hybridized offspring have rapidly dispersed 
throughout tropical South and Central America and are now as far north 
as Honduras and El Salvador (Rinderer 1986). The dispersion from their 
original importation site into new areas has been rapid — 200-500 km per 
year (Taylor 1977, 1985; Winston 1979a). As Africanized honey bees have 
colonized new areas, they have achieved dramatic population densities 
(Michener 1975). There are now probably more than ten million feral 
colonies in South and Central America (Winston, Taylor and Otis 1983). 
Their success in these new habitats, compared with the lack of success 
of European honey bee populations, may be attributed to their foraging 
behavior which is more suited to the resource patterns of the tropics 
(Nunez 1973, 1979a, 1982; Rinderer, Bolten, Collins and Harbo 1984; 



3 

Rinderer, Collins and Tucker 1985; Winston and Katz 1982). As a result 
of both foraging success and the length of time throughout the year that 
resources are available in the tropics. Africanized honey bees have a 
high annual reproductive rate* which is responsible for both their rate 
of dispersal into new areas and their high colony densities. Net 
reproductive rates for Africanized bees are estimated to be 15 colonies 
per colony per year based on demographic data collected in French Guiana 
(Otis 1980, 1982a), compared with 0.92-0.96 (Seeley 1978) or, when 
afterswarms are considered, 3-3.6 (Winston 1980a; Winston, Taylor and 
Otis 1983) colonies per colony per year for European honey bees in North 
America. 

Characteristics of Africanized Honey Bees 
The most well known characteristic that differentiates Africanized 
honey bees from European honey bees is their stinging behavior (Collins, 
Rinderer, Harbo and Bolten 1982; Stort 1974, 1975a, 1975b, 1975c, 1976). 
Because of their stinging behavior. Africanized bees are a health hazard 
for both humans and domestic animals (Taylor 1986). Collins, Rinderer, 
Harbo and Bolten (1982) compared the colony defense behavior of 
Africanized honey bees in Venezuela with European bees under identical 
conditions in Venezuela and with a population of European bees in 
Louisiana, U.S.A. Africanized honey bees responded more rapidly and in 
much greater numbers, resulting in 5.9 times as many stings in a target 
compared with European honey bees in Venezuela and 8.2 times as many 
stings compared with European bees in Louisiana. Two additional 
components of Africanized honey bee defense behavior increase their 
potential as a health hazard. Compared to European bees. Africanized 
bees pursue a source of disturbance for a greater distance (160 versus 



4 
22 meters) and remain disturbed for a greater period of time (28 versus 
3 minutes) (Stort 1971 cited in Goncalves 1974). Differences in defense 
behavior between Africanized and European bees do not appear to be a 
function of either quantitative differences in pheromone production 
(Crewe and Hastings 1976) or numbers of olfactory structures on the 
antennae (Stort and Barelli 1981). 

There is a difference in natural comb cell size between Africanized 
and European populations. The width between opposite sides of the 
hexagonal cells for the Africanized population in Brazil averaged 5.0 mm 
compared with 5.4 mm for the European population in Canada (Michener 
1972). In a recent study, cells built by Africanized swarms in 
Venezuela were 4.8-4.9 mm wide and those built by European swarms in 
Louisiana, U.S.A. were 5.2-5.3 mm wide (Rinderer, Tucker and Collins 
1982). Adult Africanized bees are smaller than European bees (62 mg 
compared with 93 mg, unengorged) (Otis 1982b; Otis, Winston and Taylor 
1981). However, as Africanized honey bees disperse into areas with 
extensive European honey bee populations, size differences between the 
two populations may become less distinct. Increased hybridization 
between the two populations could result in bees with an Africanized 
genome developing in European comb cells, resulting in larger 
Africanized bees. Therefore, methods used to identify Africanized honey 
bees based on size parameters, for example, morphometric analysis, may 
become less reliable. As Daly, Hoelmer, Norman and Allen (1982) point 
out, there is a "difficulty in using phenotype characters to identify 
genetically different, but closely related populations" (p. 593). This 
will be more evident as Africanized honey bees disperse into areas of 
Central America and particularly Mexico, where large populations of 



5 
European honey bees exist. Factors determining honey bee size and 
potential problems of Africanized honey bee identification based on size 
are analyzed in Chapter III. 

The cuticular hydrocarbon composition of Africanized honey bees is 
significantly different from that of European honey bees (Carlson and 
Bolten 1984). The differences are particularly striking for the 35, 37, 
39, 41 and 43 carbon alkenes and alkadienes that total over 22% of the 
hydrocarbons extracted from Africanized bees but only 1-3% of the 
hydrocarbons extracted from European bees. Because hydrocarbon 
composition is not affected by honey bee size or diet, using hydrocarbon 
analysis to distinguish between Africanized and European honey bees has 
great potential. However, further research to determine heritability 
patterns for different hydrocarbon components is needed. 

Differences between Africanized and European honey bees have also 
been demonstrated for foraging behavior (Nunez 1973, 1979a, 1982; 
Rinderer, Bolten, Collins and Harbo 1984; Rinderer, Collins and Tucker 
1985; Winston and Katz 1982), egg development times (Harbo, Bolten, 
Rinderer and Collins 1981), selection preferences for nest cavity sizes 
(Fletcher 1976; Michener 1972; Rinderer, Collins, Bolten and Harbo 1981; 
Rinderer, Tucker and Collins 1982), hoarding behavior (Rinderer, Bolten, 
Harbo and Collins 1982), worker bee longevity (Winston and Katz 1981), 
morphometric analysis (Daly and Balling 1978), and allozyme patterns 
(Nunamaker and Wilson 1981; Sylvester 1982). Africanized honey bee 
populations in South America are reported to have a high colony 
reproductive (swarming) rate compared with European honey bee 
populations in North America (Otis 1980, 1982a; Winston 1979b, 1980a; 
Winston, Dropkin and Taylor 1981; Winston, Taylor and Otis 1983). 



6 
However* those investigations have not been conducted under similar 
environmental or experimental conditions. Therefore, comparisons of 
reproductive rates between Africanized and European honey bees using 
those data are inappropriate for either identifying differences in 
reproductive rates for tropical and temperate honey bee populations or 
for identifying factors responsible for the success of Africanized bees 
in tropical regions. 

Purpose of My Research 

African and European honey bee populations evolved under very 
different resource and climatic conditions. The presence of both 
Africanized and European honey bees in Venezuela provided the 
opportunity to study both populations under identical conditions in the 
tropics. Differences between the two honey bee populations that make 
Africanized bees more successful in tropical regions could then be 
evaluated. The underlying assumption of my research was that the life 
history of Africanized honey bee populations in South America (as well 
as the parental population in Africa) is characterized by a high 
reproductive rate. Demographic features expected to be correlated with 
this high rate of colony reproduction include short worker bee 
development times, small worker bee size, rapid queen development and 
maturation, and increased egg laying and brood production rates. 
Predictions involving these demographic characteristics led to a series 
of experiments that are presented and discussed in the following 
chapters. 

In addition, the question of reproductive isolation versus 
hybridization and differential selection between the two populations in 
tropical conditions was experimentally evaluated. Whether there is 



7 

hybridization or reproductive isolation between Africanized and European 
honey bee populations could result in very different scenarios for the 
potential impact of Africanized honey bees on North America, 
particularly the U.S.A. 

Identification of Honev B ees Used in Mv Research 
For the experiments presented here. Africanized honey bee colonies 
were established from queens removed from feral colonies in an area in 
eastern Venezuela where there were no known European honey bees. They 
were identified as Africanized bees primarily by their distinctly 
smaller comb cell size as compared with European honey bees. 

European honey bees used in the experiments were from commercially 
produced queens from three different queen breeders in the U.S.A. 
Additional European lines were obtained from the U.S. Department of 
Agriculture Bee Research Laboratories in Madison, Wisconsin, and Baton 
Rouge, Louisiana. All of these European queens were either naturally 
mated or instrumentally inseminated in the U.S.A. and then shipped to 
Venezuela. 

When and Whe re Research Was Conducted 
All field research with Africanized and European honey bees was 
conducted from December 1978 through February 1980 at the Ministerio de 
Agricultura y Cria de Venezuela Africanized Honey Bee Research 
facilities near Maturin, Monagas. The area originally was a Tropical 
Dry Forest Csensu Holdridge Life Zone System (Holdridge 1964; Ewel and 
Madriz 1968)]. The forest had been partially cleared, and the area was 
grazed by cattle. 



8 

All Africanized and European honey bee comparisons were made at the 
same time under identical experimental conditions. Field and 
experimental methods are described for each of the experiments in the 
appropriate chapters. 

A few experiments with European honey bees only were conducted in 
the U.S.A. to confirm techniques developed and used while in Venezuela. 
These experiments were undertaken either at the U.S. Department of 
Agriculture Bee Breeding and Stock Center Laboratory in Baton Rouge, 
Louisiana, or at the bee research facilities of the Institute of Food 
and Agricultural Sciences at the University of Florida, Gainesville. 



CHAPTER II 
WORKER BEE DEVELOPMEhfT TIMES 



Introduction 

The presence of both Africanized and European honey beeSf Apis 
me11 ifera » in South America allows for comparisons to be made under 
identical conditions between a population that has evolved in the 
tropics and one that has evolved in temperate regions. Africanized 
honey bee populations in South America are reported to have a high 
colony reproductive (swarming) rate compared with European honey bee 
populations in North America (Otis 1980, 1982a; Seeley 1978; Winston 
1979b, 1980a; Winston, Dropkin and Taylor 1981; Winston, Taylor and Otis 
1983). However, these investigations have not been conducted under 
similar environmental or experimental conditions. Therefore, 
comparisons of reproductive rates between Africanized and European honey 
bees using these data are inappropriate either for identifying 
differences in reproductive rates between tropical and temperate honey 
bee populations or for identifying factors responsible for the success 
of Africanized bees in tropical regions. 

Reproductive rates in honey bees are a function of colony growth 
rates which are a result of an interaction of at least three factors: 
resource availability, resource utilization efficiency (foraging 
success, brood production efficiency, and bee size), and colony 
demographic parameters. Worker bee longevity is the only demographic 



10 
parameter that has been compared between Africanized and European bees 
under identical conditions. The greater longevity of European honey 
bees (Winston and Katz 1981) gives European bees a colony growth rate 
advantage. Other demographic characteristics that affect reproductive 
rates of Africanized and European honey bees (for example, worker bee 
development times, brood mortality, queen development and maturation 
periods, queen fecundity and brood production rates) have not been 
evaluated for Africanized and European bees under similar conditions. 
As part of a larger study evaluating these demographic parameters, this 
study compares worker bee development periods for Africanized and 
European honey bees in Venezuela. 

Smith (1958a) and Tribe and Fletcher (1977) reported that the total 
development period (from oviposit ion to adult emergence) for worker bees 
of Apis mellifera adansonii (now classified as h, HL. scutellata; Ruttner 
1976a, 1976b, 1981) from South Africa was between 18.6-20 days. Similar 
development times for the Africanized honey bee populations (descendents 
of A. rn.. scutellata ) in Brazil have been presented (Kerr, Goncalves, 
Blotta and Maciel 1972; Wiese 1972). Worker bee development times for 
European populations (primarily A. m. mellifera , 1 igustica , carnica and 
caucasica ) from Europe and North America range from 20-24 days (Jay 
1963 ) . 

The differences in development times between African (and 
Africanized) and European genotypes, which range from 1.4 to 5.4 days, 
are difficult to evaluate because they are based on data collected under 
very different experimental conditions. Jay (1963) summarized a number 
of factors that affect development times: seasonal variation in 
temperature; temperature differences in different areas within the brood 



11 

nest; colony size, which affects both brood nest temperature and feeding 
frequency and quality; and nectar and pollen resources, which also 
affect feeding quantity and quality. Valid comparison of development 
times between genotypes or populations can only be made when these 
factors are controlled under similar experimental conditions. 

The importance of slight temperature differences on development 
time cannot be overstated. Development times for European worker bees 
in Wisconsin, U.S.A., averaged 20.5 days but ranged from 20-24 days, 
depending on differences in temperature in different areas of the brood 
nest (Milum 1930). Harbo and Bolten (1981) showed that fertilized eggs 
kept at 34.8°C hatched about 1.4 hours sooner than those kept at 34.3°C. 
This difference in egg hatch time for only a 0.5°C difference in 
temperature can be extrapolated to approximately 10 hours for the entire 
development period [calculated from Harbo and Bolten (1981)]. However, 
normal temperatures within a brood nest can vary to a much greater 
extent (Milum 1930; Jay 1963). For example, when Tribe and Fletcher 
(1977) determined the development rates of African honey bees in South 
Africa, they recorded that temperatures in the brood nest varied from 
26-3 4°C. 

In an incubator with controlled temperature and humidity, eggs from 
Africanized genotypes hatched significantly sooner than eggs from 
European genotypes, 69.6 +1.06 hours compared with 73.3 +1.14 hours 
(Harbo, Bolten, Rinderer and Collins 1981). Egg development requires 
that only temperature and humidity be controlled and can therefore be 
evaluated independently of colony-level parameters. However, 
differences between Africanized and European genotypes for total worker 
bee development periods need to be evaluated within a colony in order to 



12 
allow for normal feeding and growth. Worker bee development rates are a 
result of an interaction between the egg genotype and the colony. There 
are three colony-level factors that need to be considered when comparing 
total development time of Africanized and European worker bees. 

First is the effect of comb cell size. There is a difference in 
natural comb cell size between Africanized and European populations. 
The width between opposite sides of the hexagonal cells for the African 
population in Africa measured 4.77-4.94 mm (Smith 1958a). Cells for the 
Africanized population in Brazil averaged 5.0 mm (range 4.8-5.4 mm) 
(Michener 1972), but cells of the Africanized population in Venezuela 
averaged 4.8 mm (range 4.5-5.0 mm) (Chapter III; Rinderer, Tucker and 
Collins 1982). Cells from the European population from Ontario, Canada, 
averaged 5.4 mm (range 5.2-5.7 mm) (Michener 1972), and those from 
Louisiana, U.S.A., averaged 5.2-5.3 mm (range 5.2-5.4 mm) (Rinderer, 
Tucker and Collins 1982). Adult bee size is a function of comb cell 
size (Grout 1937); adult Africanized bees are smaller than European bees 
(62 mg compared with 93 mg, unengorged) (Otis 1982b; Otis, Winston and 
Taylor 1981). 

Abdellatif (1965) suggested that larvae in smaller comb cells 
received less food which caused them to elongate and become sealed 
earlier. Also, Tribe and Fletcher (1977) suggested that the difference 
in development time for African and European genotypes may be a function 
of the small African bee size. Therefore, the effect of comb cell size 
needs to be considered when comparing development times of Africanized 
and European honey bees. 

The second colony-level factor is the effect of nurse bee genotype. 
There may be behavioral differences and/or physiological differences in 



13 
the way 1n which nurse bees from the two populations interact with the 
developing larvae. Mel'nichenko (1962) suggested that differences 
between nurse bee genotypes might affect developmental rates as well as 
size of developing larvae. For European honey bees, Lindauer (1953) 
calculated that each developing larva requires over 2785 adult bee 
visits taking a total of 10.3 hours. This appears to provide sufficient 
opportunity for possible genotype differences, either quantitative or 
qualitative, to affect development rates. In addition to potential 
qualitative or quantitative differences in feeding of larvae, nurse bees 
of different genotypes may also maintain different brood nest 
temperatures. Therefore, development times for Africanized and European 
worker bees were evaluated in both Africanized and European colonies. 

The third colony-level factor affecting worker development is 
colony size (number of worker bees in a colony). Colony size affects 
both brood nest temperature and larval feeding rates, which, as already 
discussed, are two major factors affecting development times. 

In addition to these colony-level parameters, resource conditions 
also affect development time. Nelson and Sturtevant (1924) reported 
that development of European bees was more rapid with increased larval 
feeding associated with a nectar flow. Ribbands (1953) and Jay (1963) 
both summarized evidence of the effect of food on larval development 
rates. Therefore, all comparisons of worker development times were 
conducted simultaneously to avoid any differences due to resource 
conditions. 

This paper reports the results from a comparison of the development 
times of Africanized and European honey bees under identical conditions 
in Venezuela. The experimental design allowed for the discrimination 



14 \ 



between the effects of egg genotype and the colony-level parameters of 
comb cell size and nurse bee genotype on worker bee development time. 
These experiments were conducted during July-October 1979. 

Methods 
Table 2-1 summarizes the experimental design. Four experimental 
colony treatments were established as follows: 

i. Africanized comb cell size. Africanized nurse bees (ASS) 
ii. Africanized comb cell size, European nurse bees (H2) 
iii. European comb cell size, Africanized nurse bees (A41) 
iv. European comb cell size, European nurse bees (IBR877). 
Each experimental colony was a five-frame hive (22 liters) with 
four empty combs and one comb with honey and pollen and approximately 2 
kg of young adult bees (Africanized or European, depending upon 
treatment). Because natural nectar and pollen resources were available 
irregularly throughout the experimental period (16 weeks), the colonies 
were supplemented with honey and pollen as necessary. 

European comb was built from commercially-produced beeswax 
foundation that had been fastened into standard wooden frames. 
Africanized comb was naturally built (not from foundation) by 
Africanized bees in empty standard wooden frames to facilitate 
manipulation and colony inspection. 

The queens in colonies with Africanized nurse bees (ASS and A41) 
were Africanized queens produced by standard queen rearing methods 
(Laidlaw 1979) and then naturally mated to Africanized drones. Mating 
occurred in an area of eastern Venezuela that had a large feral 
population of Africanized honey bees with no known European honey bees 
present (near San Jose de Buja, Monagas, Venezuela). The feral colonies 



15 
from which the Africanized queen mothers were extracted were also from 
this area. The colonies were identified as Africanized honey bees by 
both their behavior and their small comb cell size characteristic of the 
Africanized population (4.5-5.0 mm, see Chapter III). 

The queens in the colonies with European nurse bees (H2 and IBR877) 
were European queens that had been mated to European drones in the 
U.S.A. and transported to Venezuela. Line H2 was from a commerical 
queen producer in the southeastern U.S.A.; IBR877 was an outbreed line 
from the U.S. Department of Agriculture Bee Breeding and Stock Center 
Laboratory in Baton Rouge, Louisiana, U.S.A. 

The source for the Africanized egg genotype (A26) was a queen 
removed from a feral colony of Africanized honey bees in the San Jose de 
Buja area. The colony was identified as Africanized by its behavior and 
characteristic comb cell size. The source of the European egg genotype 
(Y5) was a queen commercially produced in the southeastern U.S.A. and 
shipped to Venezuela. 

Because adult longevity is from 2 to 5 weeks for European honey 
bees (Woyke 1984) and 2 to 3 weeks (or less) for Africanized honey bees 
(Winston 1979bj Winston and Katz 1981), experimental colonies were 
established 10 weeks prior to the start of the experiment. This was 
sufficient time to insure that at the beginning of the developmental 
trial all the adult bees present within the experimental colonies had 
developed in those colonies, and, therefore, were offspring of a known 
genetic line having developed within a known comb cell size. Before 
development times were measured, the worker bee populations in the 
experimental colonies were equalized as much as possible by removing 
random samples of bees from the most populous colonies. 



16 
The experimental colonies were placed in an apiary under a roof 
with completely open sides. The roof served two purposes. First, the 
colonies were in complete shade, which reduced any effects from 
differences in ambient temperature and sunlight. Second, colonies could 
be opened and inspected in order to monitor development with a minimum 
of disturbance, especially during rain. 

Eggs were collected from queens of the two designated egg source 
lines (A26 and Y5), that were established in modified colonies similar 
to those used by queen producers in the U.S.A. (Harp 1973). These 
colonies consisted of five standard frames with the middle frame 
isolated from the four others by queen excluder side and top panels. 
The excluder panels have a mesh size that restricts the queen from 
passing through because of her wider thorax, but allows worker bees to 
pass through to feed and communicate with the queen. Thus, the queen 
was isolated on a specific comb so that eggs could be collected that 
would then be placed into one of the four experimental colonies to 
evaluate development time. 

The comb used had the appropriate comb cell size for the 
experimental colony into which it would be placed. Comb cell size was 
measured in each experimental nurse bee colony and for each Africanized 
and European egg comb put into each experimental colony. 

The queens were caged on each comb for 24 hours so that a large, 
uniform egg sample could be collected. Eggs were monitored only from 
the center of each frame, which insured a more uniform temperature 
during development as well as uniform brood nest position. The large 
egg sample also insured that the monitored eggs were in a normal 
environment, surrounded by similarly-aged developing bees. The thirty 



17 
or forty eggs selected to be monitored for development formed either a 
3x10 or a 4x10 cell area. A reference point which facilitated locating 
the designated development sample was indicated by colored pins inserted 
five cells to the left of each original egg row. 

The two combs (one comb from each egg source) were placed in the 
center of each experimental colony at the same time. The cells with the 
eggs to be monitored faced each other in order to reduce any effects of 
brood nest position. Two frames of brood were removed from each 
experimental colony to make room for the experimental frames. This also 
reduced the amount of brood being reared in each experimental colony, 
insuring that the monitored eggs would be optimally fed. 

The pairs of frames were put into the four experimental colonies on 
four successive days because of the 24 hours needed to collect each set 
of eggs. Once the eggs were put into the experimental colonies, they 
were inspected every day at 0800 hours. The survivorship and 
developmental status of each original test group egg was recorded. 

The sample size of eggs monitored was selected to minimize the time 
each colony would need to be opened for observation in order to minimize 
disturbance. When a colony was disturbed, bees flew from the comb and 
ran on the bottom of the hive, resulting in temperature fluctuations and 
interrupted feeding of larvae. Colonies were carefully opened, using 
minimal amounts of smoke. Adult bees were not shaken off the combs but 
rather gently pushed aside to observe the development stage within the 
cells. Inspections more frequent than every 24 hours also increased the 
level of disturbance, especially in the Africanized colonies (ASS and 
A41). The advantage of more frequent monitoring to more accurately 



18 
record time of developmental changes was outweighed by the negative 
effects of disturbance on development rates. 

Temperature of the brood nest In the space between the two 
monitored combs was recorded periodically ^y placing a thermometer Into 
that area through a hole In the hive cover. This was only an 
approximate measure of temperature because development occurs within the 
cells where temperature Is less affected by the ventilating air currents 
within the hive. 

Results 
Table 2-2 presents the development times for unsealed brood, sealed 
brood, and total development times from oviposltlon to adult emergence 
for Africanized and European egg genotypes In each of the four 
experimental treatments. Table 2-3 summarizes the argumentation for 
evaluating the Interactions of egg genotypes with the colony-level 
parameters of comb cell size and nurse bee genotype on worker bee 
development times. The best tests to use to compare differences between 
the Africanized and European egg genotypes are evaluating pairs In each 
of the four experimental nurse bee colonies (A55, H2, A41, and IBR877), 
I.e., AxB, CxD, ExF, and GxH. In these comparisons, the colony-level 
parameters (colony population, position within the brood nest, 
temperature, comb cell size, and nurse bee genotype) are Identical and 
allow for only differences in egg genotype to be compared. Tables 2-4, 
2-5 and 2-6 summarize the results from the statistical analyses for the 
unsealed brood, sealed brood, and total development times, respectively. 
Data were analyzed with the Kolmogorov-Smlrnov one-tailed test using the 
chl square distribution, df = 2 (Slegel 1956). 



19 
Africanized worker bees developed faster than European bees (ACEG x 
BDFH). The unsealed larval period was 4.3 +0.4 days compared with 4.9 
+0.4 days, P<0.001; the sealed larval and pupal period was 11.6 +0.5 
days compared with 11.9 +0.4 days, P<0.01; and the total development 
time was 18.9 +0.3 days compared with 19.8 +0.4 days, P<0.001. There 
was no significant effect of comb cell size or nurse bee genotype on 
development times. These differences in total development time are 
similar to the differences found between three different lines of 
Africanized and three different lines of European honey bees compared in 
another study (19.2 days compared with 20.0 days. Table A-1). 

Comb cell sizes for the experimental colonies and egg sample frames 
are presented in Table 2-7. Temperatures recorded for all experimental 
colonies varied from 35-36°C. 

When differences in development times between Africanized and 
European egg genotypes were compared for each stage of development, the 
greatest difference was observed in the unsealed larval stage (Table 2- 
8). However, this was not a result of a differential acceleration of 
development during the unsealed larval period for the Africanized honey 
bees. The proportion of unsealed larval development time to total 
development time and the proportion of sealed brood development time to 
total development time were compared for the Africanized and European 
honey bee populations following angular transformations of the 
proportions (Sokal and Rohlf 1969). These proportions were not 
significantly different between the Africanized and European honey bees. 

The differences recorded for the unsealed brood stage between the 
Africanized and European honey bees may be an artifact of the experiment 
for two reasons. First, the 24-hour observation interval may obscure 



20 
exact timing of developmental changes. Second, the process of sealing 
is not a precise developmental stage and may take from six hours 
(Lindauer 1953) to 24 hours (Jay 1963). When the unsealed and sealed 
brood stages are combined, the proportional differences between the two 
populations are the same as for egg development times and total 
development times (Table 2-8). 

Mortality for different developmental stages for each colony 
treatment is presented in Table 2-9 (see also Table A-2). Mortality was 
high (26-37%) for larvae in the experimental colony (H2) with European 
nurse bees on Africanized comb cell size. The high mortality during the 
larval stage may be a result of the reduced ability of larger European 
nurse bees to feed the developing larvae in smaller Africanized comb 
cells. There was also a high egg mortality recorded for European eggs 
in A41 and IBR877— 34 and 70%, respectively. Woyke (1977) reports 
J normal mortality may be as high as 10-50% depending on the season. 

Garofalo (1977) also reports varying mortalities depending on both the 
size of the colony and the time of year: eggs 10-25%, larvae 11-37%, 
pupae 5-7%, and all developmental stages combined 25-53%. 

Discussion 
This study is the first to evaluate worker bee development times 
between Africanized and European honey bees as an interaction between 
egg genotype and colony-level parameters. Differences in worker bee 

development times were independent of the colony-level parameters of 

,1 

comb cell size and nurse bee genotype but were dependent on egg genotype 

J 

differences between the Africanized and European populations. The 
difference in development times between these two populations was not as 



21 
large as expected from previous reports, which underscores the 
importance of making comparisons under identical conditions. 

The proportional difference (5.7%) in egg development times between 
Africanized (A26) and European (Y5) honey bees reported by Harbo, 
Bolten, Rinderer and Collins (1981) is identical to the proportional 
difference in total development time reported in the present study 
(5.7%, see Table 2-8). Egg development time is a function of the 
inherent rate characteristic of the particular genotype because colony- 
level parameters (e.g., feeding) are not involved (Harbo, Bolten, 
Rinderer and Collins 1981). Using egg development to evaluate 
differences in total development between genotypes (or populations) is 
advantageous because egg development times are easier to evaluate, take 
less time, have fewer variables to control (temperature and humidity 
only), and can be evaluated in an incubator rather than in a colony, 
avoiding problems associated with disturbing the colony during 
observations. It must be noted, however, that by using egg development 
times one can only extrapolate proportional differences between 
genotypes for total development time but cannot extrapolate the absolute 
total development time. 

A prerequisite for high reproductive rates would be a rapid colony 
growth rate. However, the importance of worker development time to the 
rate of colony growth (increase in numbers of bees in a colony) has 
apparently been misunderstood, e.g., see Fletcher (1977a, 1978), 
Fletcher and Tribe (1977a), Tribe and Fletcher (1977), Winston (1979b), 
Winston, Dropkin and Taylor (1981), Winston and Katz (1982), Winston, 
Taylor and Otis (1983). The difference in worker development times 
observed for Africanized and European honey bees is not a factor 



22 

contributing to either differences in rate of colony population increase 
or to differences in reproductive rates between the two honey bee 
populations. 

The importance attributed to worker development time on the rate of 
colony growth may be a result of confusing colony population increase 
(increase in the number of bees in the colony) with general population 
growth models designed for other species in which all individuals are 
potential reproductives. For honey bees, individual (or worker bee) 
development time is not equal to generation time. Organism growth 
models must be used to evaluate colony growth even though the number of 
individual worker bees within the hive increases. The hive is the 
organism. Worker bee development time does not affect the rate of 
colony growth. Worker development time affects only the length of time 
between a given change in egg laying rate and its resulting change in 
population increase or decrease. Africanized bees develop in 19 days 
and begin their population increase (=growth) on the 19th day of the 
colony cycle, compared with the 20th day for European bees. This 
difference is trivial compared to potential differences from other 
demographic factors that do affect rates of colony growth. Egg laying 
and brood production rates, worker bee longevity, brood mortality, and 
resource availability are factors that do affect the rate of colony 
population increase and, therefore, affect the reproductive rates. 

Tribe and Fletcher (1977) have suggested that African worker bees 
have a shorter unsealed development stage because they do not grow as 
large as European honey bees. They compare their data for African bees 
with data for European bees in the literature and conclude that African 
bees have a 20-30% shorter unsealed larval stage. There are four 



23 

problems with their analysis. First, as already pointed out, using the 
duration of the unsealed stage has inherent problems because it is not a 
precise development stage. Second, comparisons based on data collected 
under different experimental conditions are not valid. Third, their 
logic is perhaps circular with respect to the question of larval size 
and larval development times. In the present study, development time 
was not size-related for either Africanized or European honey bees. 
Africanized honey bees that developed in European comb had the same 
development times as those that developed in Africanized comb even 
though Africanized bees reared in European comb were significantly 
larger (16%; Chapter III). The same relationship was true for European 
honey bees with a 17% increase in size of bees from European comb 
compared with bees from Africanized comb. And fourth, their comparison 
is in itself incorrect. Rather than compare the differences in unsealed 
development times between African and European populations to determine 
if the African population has a relatively shorter duration as unsealed 
larvae, they should have used the proportion of unsealed development 
period to total development period in order to compare African and 
European populations. In the present study, the relative times spent as 
an unsealed larvae to the total development time for both the 
Africanized and European genotypes were not significantly different. 
The differences in development time between Africanized and European 
populations appear constant throughout development without any 
developmental acceleration during the larval stage for either 
Africanized or European honey bees. 



24 



TABLE 2-1. Experimental matrix for evaluating interaction of egg 

genotype, comb cell size, and nurse bee genotype on worker 
development times. A - H represent each treatment. 



AFRICANIZED EUROPEAN 

EGG GENOTYPE (A26) EGG GENOTYPE (Y5) 



AFRICANIZED COMB CELLS 

AFRICANIZED NURSE 

BEES (ASS) A B 

EUROPEAN NURSE 

BEES (H2) C 



EUROPEAN COMB CELLS 

AFRICANIZED NURSE 
BEES (A41) 

EUROPEAN NURSE 
BEES (IBR877) 



25 



TABLE 2-2. Interaction of egg genotype, comb cell size, and nurse 
bee genotype on worker bee development time (days): 
median, (range), mean + SD, (n = sample size). 



AFRICANIZED EGG GENOTYPE (A26) 



us^ sb"^ TDT 



AFRICANIZED COMB CELLS 





4.0 


12.0 


19.0 


AFRICANIZED NURSE 


(4-5) 


(11-12) 


(18-20) 


BEES (A55) 


4.2 +0.4 


11.6 +0.5 


18.8 +0.5 




(n = 30) 


(n = 30) 


(n = 30) 




4.0 


12.0 


19.0 


EUROPEAN NURSE 


(4-5) 


(11-12) 


(19-20) 


BEES (H2) 


4.4 + 0.5 


11.6 +0.5 


19.1 +0.2 




(n = 29) 


(n = 29) 


(n = 29) 



EUROPEAN COMB CELLS 



AFRICANIZED NURSE 
BEES (A41) 


4.0 
(4-5) 
4.2 +0.4 
(n = 30) 


12.0 
(11-12) 
11.7 +0.5 
(n = 30) 


19.0 

(18-19) 

18.9 +0.2 

(n = 30) 


EUROPEAN NURSE 
BEES (IBR877) 


4.0 

(4-5) 

4.3 + 0.5 

(n = 26) 


12.0 
(11-12) 
11.7 +0.5 
(n = 26) 


19.0 
(19) 
19.0 +0 
(n = 26) 


TOTALS 


4.0 

(4-5) 

4.3 + 0.4 

(n = 115) 


12.0 
(11-12) 
11.6 +0.5 
(n = 115) 


19.0 
(18-20) 
18.9 + 0.3 
(n = 115) 



®US = unsealed brood (unsealed larval development period only). 

'^SB = sealed brood (pre-pupae and pupae). 

^TDT = total development time (oviposition to adult emergence). 



26 



TABLE 2-2— extended . 



EUROPEAN EGG GENOTTPE (Y5) 



US 


SB 


TDT 


5.0 


12.0 


20.0 


(4-5) 


(11-12) 


(19-20) 


4.9 1 0.3 


11.8 + 0.4 


19.6 + 0.5 


(n = 37) 


(n = 37) 


(n = 37) 


5.0 


12.0 


20.0 


(4-5) 


(12) 


(19-20) 


4.9 +0.4 


12.0 + 


19.9 + 0.4 


(n = 22) 


(n = 22) 


(n = 22) 



5.0 12.0 20.0 

(4-6) (12-13) (19-21) 

5.0 +0.4 12.0 +0.2 20.0 +0.3 

(n = 19) (n = 19) (n = 19) 



5.0 12.0 20.0 

(4-6) (12-13) (19-21) 

4.7+0.8 12.1+0.4 19.8+0.7 

(n =7) (n = 7) (n = 7) 



5.0 12.0 20.0 

(4-6) (11-13) (19-21) 

4.9 +0.4 11.9 +0.4 19.8 +0.4 

(n = 85) (n = 85) (n = 85) 



27 

TABLE 2-3. Summary of hypotheses and tests for evaluating development 
times; letters represent treatments (see Table 2-1). 

HI: Worker bee development is faster for Africanized genotypes than 
for European genotypes. 

/^ X ^ African'ized comb cell size; Africanized nurse bees 

^ X ^ Africanized comb cell size; European nurse bees 

^ ^ f^ European comb cell size; Africanized nurse bees 

G X ^ European comb cell size; European nurse bees 

^ ><■ ^ Africanized comb cell size and nurse bees compared with 

European comb cell size and nurse bees 

AC X BD Africanized comb cell size; both nurse bee genotypes 
combined 

^^ ^ ^^ European comb cell size; both nurse bee genotypes combined 

AE X BF Africanized nurse bees; both comb cell sizes combined 

^^ X DH European nurse bees; both comb cell sizes combined 

ACEG X BDFH Both comb cell size and both nurse bee genotype variables 
combined 



H2: Worker bee development is more rapid in Africanized comb cells 
than in European comb cells. 

^ ^ E Africanized egg genotype; Africanized nurse bees 

C X G Africanized egg genotype; European nurse bees 

B X F European egg genotype; Africanized nurse bees 

D X H European egg genotype; European nurse bees 

^^ X EG Africanized egg genotype; both nurse bee genotypes 
combined 

BD X FH European egg genotype; both nurse bee genotypes combined 

H3: Worker bee development is more rapid with Africanized nurse bees 
than with European nurse bees. 

^ X- ^ Africanized egg genotype; Africanized comb cell size 

E X G Africanized egg genotype; European comb cell size 

^ ^ ^ European egg genotype; Africanized comb cell size 

F X H European egg genotype; European comb cell size 

AE X CG Africanized egg genotype; both comb cell sizes combined 

BF X DH European egg genotype; both comb cell sizes combined 



28 

TABLE 2-^ — continiiAd . 



H4: Worker bee development 1s more rapid with Africanized comb cells 
and Africanized nurse bees than with European comb cells and 
European nurse bees. 



A X G Africanized egg genotype 
^ X H European egg genotype 



29 

TABLE 2-4. Unsealed brood development times. Hypotheses were tested 
using Kolmogorov-Smirnov one-tailed test* chi-square 
distribution, df = 2, alpha = 0.05 (Siegel 1956). 

HI: Worker bee development is faster for Africanized genotypes than for 
European genotypes. 

A X B ***^ 
C X D ** 

E X F *** 
G X H NS 

A X H NS 

AC X BD **« 

EG X FH *** 

AE X BF »«» 

CG X DH *»* 

ACEG X BDFH *** 

H2: Worker bee development is more rapid in Africanized comb cells than 
in European comb cells. 

AXE NS 

C X G NS 

B X F NS 

D X H NS 

AC X EG NS 

BD X FH NS 

H3: Worker bee development is more rapid with Africanized nurse bees 
than with European nurse bees. 

A X C NS 

E X G NS 

B X D NS 

F X H NS 

AE X CG NS 

BF X DH NS 

H4: Worker bee development is more rapid with Africanized comb cells 
and Africanized nurse bees than with European comb cells and 
European nurse bees. 

A X G NS 

B X H NS 



^ ** = P<0.01 
^*** = p<o.001. 

Analysis may be NS because test used is conservative for small sample 

sizes using chi-square distribution. 



30 

TABLE 2-5. Sealed brood development times. Hypotheses were tested 
using Kolmogorov-Smirnov one-tailed test, chi-square 
distribution, df = 2, alpha = 0.05 (Siegel 1956). 

HI: Worker bee development is faster for Africanized genotypes than for 
buropean genotypes. 

A X B NS 
C X D *a 

E X F NS 

G X H Ns 

A X H Ns 
AC X BD » 

EG X FH « 

AE X BF * 

CG X DH «* 

ACEG X BDF »* 



H2: 



H4: 



Worker bee development is more rapid in Africanized comb cells than 

in EuroDfifln mml-i (~cil1e 



in European comb cells 



AXE NS 

C X G NS 

B X F US 

D X H NS 

AC X EG NS 

BD X FH NS 

H3: Worker bee development is more rapid with Africanized nurse bees 
than with European nurse bees 

A X C NS 

E X G NS 

B X D NS 

F X H NS 

AE X CG NS 

BF X DH NS 



Worker bee development is more rapid with Africanized comb cells 
and Africanized nurse bees than with European comb cells and 
European nurse bees. 



A X G NS 

B X H NS 



* = P<0.05 
** = P<0.01. 



A X B 


C X D 


E X F 


G X H 


A X H 


AC X BD 


EG X FH 


AE X BF 


CG X DH 


ACEG X BDFH 



31 

TABLE 2-6. Total worker bee development times. Hypotheses were tested 
using Kolmogorov-Smirnov one-tailed test, chi-square 
distribution, df = 2, alpha = 0.05 (Siegel 1955). 

HI: Worker bee development is faster for Africanized genotypes than for 
European genotypes. 

*** 
«w« 
»*# 
** 
#*# 
»#« 
«** 
*** 
»** 

H2: Worker bee development is more rapid in Africanized comb cells than 
in European comb cells. 

AXE NS 

C X G NS 

B X F MS 

D X H NS 

AC X EG NS 

BD X FH NS 

H3: Worker bee development is more rapid with Africanized nurse bees 
than with European nurse bees 

A X C NS 

E X G NS 

B X D NS 

F X H NS 

AE X CG NS 

BF X DH NS 

H4: Worker bee development is more rapid with Africanized comb cells 
and Africanized nurse bees than with European comb cells and 
European nurse bees. 

A X G NS 

B X H NS 



^ ** = P<0.01 
*** = P<0.001. 



32 



TABLE 2-7. Comb cell size for worker development time experiment: comb 
measurements = mm for 10 consecutive, horizontal cells, mean 
+ SD, (sample size) . 



COMB CELL SIZE 



hAJRSE BEE AFRICANIZED EGG EUROPEAN EGG 
COLONY GENOTYPE GENOTYPE 



AFRICANIZED COMB CELL SIZE^ 

AFRICANIZED NURSE 47.5+0.58 49.8+0.50 45.8+0.50 



BEES (A55) (4) (^) 



BEES (A41) (3) (f) 



Natural comb built without foundation. 
Built from foundation. 



(4) 



EUROPEAN NURSE 48.2 +0.96 48.5 +0.58 48.5 + 0.58 
BEES (H2) (4) (4) (4) 

EUROPEAN COMB CELL SIZE"^ 

AFRICANIZED NURSE 54.0 +0.0 54.0 +0.0 54.0 + 0.0 



(3) 



EUROPEAN NURSE 53.3+0.58 53.3+0.58 53.7+0.58 

BEES (IBR877) (3) (J) (f) 



33 



TABLE 2-8. 



Compartson of differences 1n development times (In days) 
for Africanized and European honey bees for different 
developmental stages. 



EGG HRS 
(DAYS)^ 



AFRICANIZED EGG 69.6 
GENDTYPE (A26) (2.90) 



DEVELOPMENTTAL STAGES 



UNSEALED SEALED 
BROOD ° BROOD^ 



4.3 



11.6 



UNSEALED TOTAL 
& SEALED DEVELOPMEhfT 



15.9 



18.80 



EUROPEAN EGG 73.6 
GENOTYPE (Y5) (3.07) 



4.9 



11.9 



16.8 



19.87 



% DIFFERENCE*^ 



5.7 



14.0 



2.6 



5.7 



5.7 



^From Harbo, Bolten, Rinderer and Collins (1981); data used are 
their Africanized #3 = A26 and their European #5 = Y5. 
Unsealed larval period only. 

^Pre-pupae and pupae. 

H Difference = [(Y5)-(A26)/(A26)] x 100. 



34 



TABLE 2-9. Mortality during different developmental stages. 



AFRICANIZED EGG GENOTYPE (A26) EUROPEAN EGG GENOTYPE (Y5) 



h^ E2^ ^l" 4^ SB^ Nf El E2 L^ l^ SB N 



AFRICANIZED COMB 
CELL SIZE 

AFRICANIZED 
NURSE BEES 



ASS 

EUROPEAN 
NURSE BEES 

H2 

EUROPEAN COMB 
CELL SIZE 

AFRICANIZED 
NURSE BEES 

A41 

EUROPEAN 
NURSE BEES 



SOS 



3 7 







40 1 2 40 



40 1 13 36 



30 8 2 29 



IBR877 



2 2 



30 13 6 1 27 



Mortality during first 24 hours in test colony (acceptance), 
portal ity between 24-72 hours (before hatching). 
Jlortality between 72-96 hours (at time of hatching), 
portal ity during older larval stages, before sealing, 
portal ity during the pupal stage. 
N = total eggs monitored. 



CHAPTER III 

INfTERACTION OF MATERNAL GENOTYPE, EGG GENOTYPE AND COMB CELL SIZE ON 

HONEY BEE WORKER SIZE AND SIZE VARIATION 



Introduction 
In the evolution of eusociality in bees (Apoidea), there is a 
considerable decrease in size variation of the workers within a colony. 
Worker size variation within a colony of primitively eusocial sweat bees 
(HaHctidae) or bumble bees (Apidae) is much greater than the size 
variation of workers within colonies of highly eusocial stingless bees 
(Meliponinae: Mel ioona and Trigona ) or honey bees (Apinae: Apis ) 
(Brian 1952; Kerr and Hebling 1964; Medler 1965; Michener 1974). For 
example, the coefficient of variation (CV) for worker weights in a 
bumble bee colony may be as high as 31-37% (calculated from Brian 1952 
for Bombus agropum. Table B-1) whereas the CV for worker weights within 
a honey bee (Api$ mellifera ) colony is only 4-7% (Table B-2). 

An effect of the reduction of size variation is that the mechanism 
for the division of labor of workers within a colony shifts from being 
size dependent to primarily age dependent (Michener 1974). In the 
primitively eusocial bumble bees, division of labor is size related 
(Brian 1952); large workers may be twice the size (linear measurements) 
of small workers within the same colony (Medler 1965). In highly 
eusocial stingless bees and honey bees, division of labor is primarily 
age dependent (Free 1965; Gary 1975; Kerr and Hebling 1964; Lindauer 
1953; Seeley 1982). In honey bees, the workers proceed through a series 

35 



36 

of age-related tasks. However, the sequence and duration of the 
different stages are flexible and depend on the needs of the colony. 

An advantage of worker size variation within bumble bee colonies 
may be efficient utilization of diverse nectar and pollen resources that 
may be size dependent. Different sized workers within a colony 
specialize on those resources that they can most efficiently exploit 
(Heinrich 1979a). However, highly eusocial bees are not at a 
disadvantage with respect to resource utilization because they have 
evolved complex communication systems that allow foragers to monitor 
changing nectar conditions and to recruit workers from the colony to a 
particular resource. Therefore, both the species characterized by 
workers of highly variable sizes and those species characterized by 
uniformly-sized workers have evolved behaviors that enhance the 
efficiency of nectar and pollen exploitation. 

The difference in intra-colony worker size variation between 
primitively eusocial and highly eusocial species of bees is so 
significant that Kerr and Hebling postulated that "some controlling 
mechanism leads to reduced variances among mature workers [Meliponinae 
and Apis], which are therefore of relatively uniform size" (1964, p. 
267). Waddington (1981) hypothesized that the evolution and maintenance 
of the complex communication systems in Apis , Trigona and Mel ipona 
depend upon uniformity of worker bee size within a colony. Differences 
in bee size may result in miscommunication because resource 
"profitability" may be size dependent. For example, a high quality 
resource for a small bee may not be a high quality resource for a larger 
bee. However, a regulatory mechanism for reduced size variation has not 
been identified. 



37 
Honey bee size and size variation Is a result of both genetic and 
environmental factors. Research has focused primarily on the extrinsic 
factors that affect bee size (e.g., comb cell size, nutrition and 
temperature). Honey bee worker sizes and honey bee comb cell sizes have 
been shown to be Inter-related: because of the manner by which comb 
cells are constructed (Darwin 1859/1958), worker body size affects the 
diameter of cells they construct (Baudoux 1933; Glushkov 1958), and 
worker bee size is correlated to the size of cells in which they are 
reared (Baudoux 1933; Buchner 1955; Glushkov 1958; Grout 1937; Michailov 
1927-28 cited in Alpatov 1929; Tuenin 1927). 

This interaction between comb cell size and egg genotype may at 
first appear to provide a mechanism for both regulating bee size and 
reducing size variation among bees within a colony. However, the comb 
cell itself can become a source of variation in bee size. Although comb 
cell size appears quite uniform, especially when first constructed, the 
cells become variable in size as the number of generations reared in 
them increases, because pupal cocoons adhere to the cell walls, reducing 
cell diameter (Abdellatif 1965; Alpatov 1929; Buchner 1955; Grout 1937). 
For example, there is a 25% reduction In cell volume between cells from 
new and old combs (Table B-3). Comb cell volume has a greater variance 
than cell diameter and is not correlated with diameter (Table B-3). 

In addition to comb cell size, there are other extrinsic factors 
that affect development and resultant bee size, e.g., quantity and 
quality of larval food, and temperature and humidity at which the larvae 
and pupae are reared (Buchner 1955; Fyg 1959; Jay 1963; Kulzhinskaya 
1956; Michailov 1927-28 cited in Alpatov 1929). These same factors not 
only affect absolute size but are sources of size variation. 



38 

The Importance of the genetic component to bee size can be inferred 
from the fact that different geographic populations of honey bees differ 
with respect to worker bee size (Alpatov 1929; Ruttner 1968, 1975, 
1976a, 1976b; Wafa, Rashad and Mazeed 1965). Because honey bee queens 
mate with many different drones (Adams, Rothman, Kerr and Paulino 1977; 
Peer 1956; Roberts 1944; Taber 1954; Taber and Wendel 1958), the genetic 
component becomes an additional factor affecting size variation. 

Africanized honey bees in Venezuela (descendents of A. n. 
spgt^lUta) were smaller and had a smaller comb cell diameter (mean 4.8 
mm between opposite sides of the hexagonal cells in the comb) compared 
with European bees in Venezuela (mean 5.4 mm) (Tables B-3 and B-4; 
Rinderer, Tucker and Collins 1982). An opportunity, therefore, existed 
to experimentally evaluate the interaction of both genotype and comb 
cell size on resultant worker bee size and size variation by studying 
both the Africanized and European honey bee populations under identical 
experimental conditions. The results from this study provide 
information not only on the proximal question involving the factors 
affecting bee size but also provide a mechanism by which size variation 
may be reduced within a honey bee colony. 

Methods 
Nine genotypes were evaluated: three Africanized, three European, 
and three F-^ reciprocal hybrids. The Africanized genotypes (A26, A57, 
and B39) were established from queens removed from feral colonies 
located in an area in eastern Venezuela with no known European honey 
bees. They were identified as Africanized honey bees by their comb cell 
sizes which were significantly smaller than European comb cell sizes 
(Tables B-3 and B-4; Michener 1972, 1975; Rinderer, Tucker and Collins 



39 
1982). The European genotypes (YD28 and WEI) were imported into 
Venezuela from the U.S. Department of Agriculture Bee Breeding and Stock 
Center Laboratory in Baton Rouge, Louisiana* U.S.A., and from a 
commercial queen producer from southeastern U.S.A., respectively. Queen 
YD28 was artificially inseminated with the spermatozoa from one drone; 
queen WEI was naturally mated. The third European genotype (SDYl) was a 
daughter from line YK produced by another commercial queen producer from 
southeastern U.S.A. and artificially inseminated in Venezuela with a 
single drone from the same commercial line. 

Two reciprocal hybrid lines were established from artificially 
reared queens (Laidlaw 1979) that were instrumentally inseminated with 
spermatozoa from single drones: Africanized queen x European drone 
(SDA12) and European queen x Africanized drone (SDYIO and SDYll). The 
Africanized queen and drone source was A26. The European queen and 
drone source was line YK. The hybrid lines were therefore genetically 
similar, but were the reciprocal of each other with respect to their 
queen and drone sources. 

Queens were produced by the standard method of transferring young 
larvae from the desired queen line into artificial queen cells which 
were then introduced into cell-producing colonies (Laidlaw 1979). 
Mature queen cells were put into an incubator (35 + 1°C) 72 hours prior 
to adult emergence. Newly emerged virgins were marked for individual 
identification and then put into individual cages and maintained in a 
strong, queenless colony for approximately one week until they were 
artificially inseminated. 

Drones for instrumental inseminations were produced by caging drone 
comb containing sealed drone pupae from the desired drone source lines. 



40 

As drones emerged, they were placed into special holding cages and 
maintained in a colony so that worker bees could feed them until they 
matured. This manipulation insured that the drones used for 
inseminations were from the desired queen lines. 

To collect eggs for the bee size experiments, queens from the nine 
genotypes were confined for five hours in their own colonies to a 
section of Africanized comb (mean cell size = 4.8 mm), using 8 x 8 cm 
push-in cages. These cages were made from 3 mm mesh hardware cloth and 
had queen excluder material soldered to the top to enable worker bees to 
pass through to tend the queen (Harbo, Bolten, Rinderer and Collins 
1981). After five hours, the queens were removed from the combs. The 8 
X 8 cm sections of comb with eggs from each queen were cut out and 
fitted in-to special frames. The nine sections were then placed in a 
strong Africanized colony (Africanized nurse bees and Africanized comb 
cell size) for development. The following day, eggs were collected in 
European combs (mean cell size = 5.4 mm) using the same procedure with 
the same nine queens except that the nine sections were put into a 
European colony (European nurse bees and European comb cell size) for 
development. Having all nine egg sources for each comb cell size 
treatment (Africanized or European) develop in the same colony 
controlled for additional variables affecting development and bee size: 
temperature and humidity, nurse bee genotypes and colony size (see 
Chapter II). 

Fresh pupal weights were compared for each of the nine genotypes 
reared in both Africanized and European comb cell sizes. Pupal weights 
were measured on the 16th day after oviposition. This age corresponds 
to the period during pupal development of least weight change (Melampy 



41 
and Wniis 1939). This was confirmed for fresh pupal weights by 
weighing a sample of pupae every 24 hours from day 11.5 post oviposition 
to 17.5 days post oviposition (Table B-5). Although Africanized bees 
develop one day faster than European bees (Chapter II), pupal weights 
can be compared because there is no significant difference in weights 
between adjacent days during this period of pupal development (Table B- 
5). 

Pupal weights were used instead of adult weights in order to reduce 
variation resulting from differences in food engorgement and/or feces 
accumulation. Pupae were carefully removed from their comb cells by 
first removing the cappings and then spreading the cell walls with a 
forceps in order that the pupae could easily be removed without 
rupturing. Weights (to 1.0 mg) were recorded using Mettler Type H4 and 
H6 balances. Comb cell diameters were determined by measuring ten 
adjacent cells; three sets of measurements were made from each comb. 

Results 
Table 3-1 presents the experimental design matrix. The interaction 
of egg genotype and comb cell size on worker bee pupal weights for each 
of the nine genotypes is summarized in Table 3-2. Table 3-3 presents 
the results of the statistical analyses. When Africanized and European 
genotypes are reared simultaneously in the same colony (same comb cell 
size, nurse bee genotype, temperature and humidity, and colony size), 
the weights of the worker bees produced are different. Africanized bee 
pupae (111.1 +7.6 mg) that developed in Africanized comb cells were 
smaller than European bee pupae (123.3 +6.3 mg) that also developed in 
Africanized comb cells (ACE x MOQ, P<0.001). When worker bees of 
European genotypes are reared in Africanized comb cells, the cells are 



42 

sealed with strongly convex cappings similar to the way cells containing 
drones are sealed in order to accomodate their larger size. Africanized 
bee pupae (123.8 +6.2 mg) that developed in European comb cells were 
smaller than European bee pupae (139.5 +5,4 mg) that also developed in 
European comb cells (BDF x NPR» P<0.001). For each of the nine 
genotypes investigated* worker bee pupae that developed in Africanized 
comb cells were smaller than pupae that developed in European comb 
cells* P<0.001. There is a 43% increase in comb cell volume between 
Africanized and European combs (Table B-3)» but the Africanized and 
European genotypes only increased in pupal weight by 11.4% and 13.1%> 
respectively (Table 3-2). These results show that both genotype and 
comb cell size affect worker bee size. 

Table 3-4 presents the results for the pupal weights of the 
reciprocal F-^ hybrids and their respective maternal lines. Data from 
only European comb cells were used in order to observe genotype effects 
without the constraint of the small Africanized comb cells on European 
genotypes. Table 3-5 summarizes the results of the statistical 
analyses. The pupal weights of the hybrids from this reciprocal F-|_ 
cross were significantly different from each other (H x J; H x L; 
P<0.001), but were the same as their respective maternal line (B x H; J 
X R; L x R). 

Discussion 
Reduction of Bee Size Variation 

Bee size is a result of not only the interaction of egg genotype 
and comb cell size but also the maternal genotype. This can be seen by 
evaluating the reciprocal hybrid crosses. The genotype component for 
bee size is not a result of "simple" inheritance because pupal weights 



43 

of genetically similar, reciprocal Fj_ hybrids are not the same. Because 
reciprocal F-^ hybrids are phenotypically different from each other, but 
I phenotypically similar to their maternal lines, maternal genotype must 
] interact with cell size and egg genotype to detennine pupal weight. 

I This is the first character in honey bees that has been shown to be 
influenced by maternal inheritance. Other genetic mechanisms cannot 
explain these results. The mechanism for maternal inheritance in worker 
size may be through egg size, which has been shown to be inherited 
(Roberts and Taber 1965; Taber and Roberts 1963). 

Alles (1961) and Mel'nichenko (1962) suggested that differences 
between nurse bee genotypes might affect size of developing larvae. 
However, McGregor (1938) found that bee size was not affected by nurse 
bee genotype. In the experiments presented in this chapter, pupae from 
each of the genotypes were reared simultaneously in the same colony for 
each comb size treatment. Therefore, differences between the pupal 
weights of Africanized and European genotypes cannot be attributed to 
either nurse bee differences, cell size, or temperature but must be a 
result of both egg and maternal genotype differences. 

The importance of maternal inheritance on bee size is that it 
reduces worker bee size variation within a colony. If maternal 
inheritance were not operating, worker bees of different sizes would be 
produced within a colony because of cell size differences and genotype 
differences. The effectiveness of maternal inheritance for reducing bee 
size variation can be demonstrated by comparing the degree of variation 
for the two parameters of bee size (comb cell volume and genotype) with 
the degree of worker bee size variation. Abdellatif (1965) showed that 



44 
when comb cell size variation increased 300%, bee size variation 
increased only 50%. 

The genetic variation of worker bees within a colony is great 
because queens mate on the average with as many as 17 drones (Adams, 
Rothman, Kerr and Paulino 1977). There is some degree of mixing of 
spermatozoa in the spermatheca resulting in spermatozoa from at least 5 
to 6 drones being used during one time interval (Page and Metcalf 1982). 
Evidence that maternal inheritance reduces size variation in genetically 
diverse worker offspring can be demonstrated by evaluating the size 
variation of offspring from single-drone and multiple-drone inseminated 
queens. The progeny of queens that were inseminated by spermatozoa from 
single drones (SDA12, SDYIO, SDYll, SDYl and YD28) were expected to be 
less variable than multiply-inseminated queens (A26, A57, B39 and WED 
because all eggs from the former queens would have been fertilized by a 
genetically identical male gamete. (Drones are haploid; all spermatozoa 
are produced by mitosis and are therefore genetically identical.) 
Evaluating the coefficient of variation (CV) for each treatment of 
genotype and comb cell size, there is no difference between the 
variation of progeny from single-drone inseminations versus those from 
multiple inseminations, as shown in Table 3-6 (Mann-Whitney U test, one- 
tailed, alpha = 0.05). 

Additional evidence of maternal inheritance reducing size variation 
in genetically heterogeneous offspring comes from analyzing the results 
of the reciprocal F^ cross. Because of the influence of maternal 
inheritance, subspecific differences in size between Africanized and 
European populations were not reflected in increased size variation of 
the hybrids compared with the parental types (Table 3-4). 



45 
Further evidence of the effectiveness of maternal inheritance 
reducing size variation can be seen by comparing the size variation 
within a colony to the size variation within a population. Alpatov 
(1929) found that within honey bee colonies, worker size variation 
i (e.g., for tongue length) was less than the variation for the local 

population of a managed apiary. For seven different apiaries in Russia, 
each apiary had an average 22.3% (range 5-42%) increased variation over 
the mean colony variation within the apiary. Although the genetic 
homogeneity of the apiaries is artificially high as a result of 
management practices of the beekeepers compared with the variation of 
natural populations of animals (Alpatov 1929), within-colony variation 
was still noticeably reduced. 

Evolution of complex communication systems in highly eusocial 
species may be responsible for selection for reduced size variation 
(Waddington 1981; Waddington, Herbst and Roubik 1986). Foragers within 
honey bee colonies have the ability to communicate information to nest 
mates about the direction, distance and "profitability" of new resources 
(von Frisch 1967) which may be interpreted correctly only if worker bees 
within the colony are the same size (Waddington 1981). Profitability of 
the resource may be size-dependent as Waddington (1981) suggested. That 
is, a high quality resource for a small bee may not be a high quality 
resource for a larger bee. 

In addition to the profitablity component, correct interpretation 
of the distance component of the honey bee waggle dance (von Frisch 
1967; Wenner 1962) may also be size-dependent. Different distance 
dialects occur not only between subspecies (Boch 1957; Gould 1982) but 
also between colonies (Esch 1978 cited in Gould 1982). There is greater 



46 

variation in individual dialects in colonies that are genetically 
heterogeneous compared with colonies that are genetically homogeneous 
(Gould 1982). Variation in bee size within a colony may accentuate 
differences in distance dialects and increase the possibility of 
miscommunication. Therefore, worker size variation within a colony of 
honey bees needs to be reduced in order for a communication system that 
recruits foragers to a particular floral resource to function correctly 
and efficiently with respect to either the profitability (Waddington 
1981; Waddington, Herbst and Roubik 1986) or distance component. 
Maternal effects operate to reduce bee size variation within a colony of 
honey bees, thereby allowing their communication system to function 
effectively. 
Africanized and Furopean H onev Bee Size Differfinra 

Several hypotheses have been suggested to explain the smaller 
worker bee size of the Africanized population. One advantage suggested 
for smaller size is more rapid development times, permitting more rapid 
colony growth resulting in increased reproductive swarming (Fletcher 
1977a; Fletcher and Tribe 1977a; Tribe and Fletcher 1977). However, 
cell size and bee size do not affect development times, and, in 
addition, worker development times do not affect colony growth rates 
(Chapter II). 

Fletcher and Tribe (1977a) and Tribe and Fletcher (1977) suggested 
that smaller bee size would permit greater numbers of worker bees to be 
reared on the same amount of food compared with larger bees. Advantages 
of increased worker numbers include frequency of reproductive swarming, 
colony defense and foraging success (Wilson 1971). Thus, smaller, 
individual bee size maximizes the use of the limited food that 



47 

characterizes the unreliable nectar availability in Africa (Tribe and 
Fletcher 1977). Smaller bee size increases the resource utilization 
efficiency of Africanized honey bees and may be a factor in the success 
and high reproductive rates of Africanized honey bees compared with 
European honey bees in tropical areas of South America (see Chapter 
VIII). 

I suggest two other hypotheses to explain the advantages of smaller 
size in the Africanized population. First, smaller size is more 
efficient with respect to dissipating heat loads in tropical habitats 
(see also Heinrich 1979b). Fletcher (1978) reports that foraging may 
stop during the hottest part of the day, which would avoid the 
disadvantages of smaller size with respect to gaining a heat load. The 
sizes of two other subspecies of honey bees in Africa support this 
hypothesis. One of the smallest subspecies in Africa, A. m.. 1 itorea . is 
found in a very hot and dry area along the coast of Kenya and Tanzania. 
One of the largest subspecies, A. m. monticola . is found at higher 
elevations and colder temperatures on Mount Kenya. 

Secondly, the advantage of smaller bee size may actually lie with 
the advantages of smaller cell size. For a given nest cavity volume, a 
larger number of worker bees can be produced if cell sizes are smaller. 
There is approximately a 25% increase in the number cells for a given 
comb area with smaller Africanized comb cells compared with larger 
European comb cells. Considering the advantages of increased worker 
numbers in a colony (Wilson 1971), the increase in worker numbers as a 
result of smaller cell size may be important, particularly if nest 
cavity volumes are limited. 



48 
Because maternal inheritance affects bee size, methods that use 
components of size to identify Africanized bees, e.g., morphometric 
analysis, may be invalid. Offspring from the cross of a European queen 
X Africanized drone (SDYIO and SDYll) are the same weight as offspring 
from European queen x European drone (SDYl) (Table 3-4). More 
importantly, the offspring from the cross of a European queen x 
Africanized drone are significantly different from offspring from 
Africanized queen x Africanized drone and Africanized queen x European 
drone matings. The European queen x Africanized drone mating represents 
the most probable scenario for initial hybridization in North America 
(see Chapter VII). That is, a virgin queen from a managed or a feral 
European colony mates with Africanized drones and produces offspring 
with a 50% Africanized genome. Analyzing the offspring using size as a 
component for identification may result in a false negative 
identification of Africanized bees. The extent of the problem would 
depend upon the degree to which particular linear measurements are 
either affected by maternal inheritance and/or are correlated with bee 
weight. As Daly, Hoelmer, Norman and Allen (1982) point out, there is a 
"difficulty in using phenotype characters to identify genetically 
different, but closely related populations" (p. 593). 



49 



TABLE 3-1. Effect of comb cell size and egg genotype 
on bee pupal weights. Experimental design 
matrix (code letters A-R used in tables of 
statistical analyses). 



COMB CELL SIZE 



EGG GENOTYPES AFRICANIZED EUROPEAN 



AFRICANIZED QUEEN X 
AFRICANIZED DRONE 



AFRICANIZED QUEEN X 
EUROPEAN DRONE 

SDAI2'' 

EUROPEAN QUEEN X 
AFRICANIZED DRONE 



EUROPEAN QUEEN X 
EUROPEAN DRONE 



la 



a 



A26^ A B 

A573 C D 



B39« E F 



SDYIO^ I J 

SDYll^ K L 



YD28'^ M N 



WEl*^ P 

SDYl^ Q R 



Natural matings, multiple inseminations. 

on. 



iiauuiai Ilia Lillys; IIIUILip 

Single drone inseminatic 



50 



TABLE 3-2. Effect of comb cell size and egg genotype on bee pupal 
weights (mg). Means + SD, (sample size) . 



COMB CELL SIZE 



EGG GENOTYPES AFRICANIZED^ EUROPEAN^ % INCREASE 



AFRICANIZED QUEEN X 








AFRICANIZED DRONE 








A26C 


105.4 + 4.5 


121.7 +5.5 


15.5 


A57^ 


(60) 
117.2 + 6.8 


(80) 
128.6 + 4.4 


9.7 


839^ 


(30) 
116.5 + 3.6 


(40) 
123.1 +6.3 


5.7 




(30) 


(40) 




COMBINED 


111.1 + 7.6 


123.8 +6.2 


11.4 




(120) 


(160) 




AFRICANIZED QUEEN X 








EUROPEAN DRONE 








SDA12<^ 


112.9 +4.4 


122.4 +4.4 


8.4 




(30) 


(40) 




EUROPEAN QUEEN X 








AFRICANIZED DRONE 








SDYIO'^ 


114.6 +3.7 


133.7 +3.3 


16.7 


SDYll^ 


(10) 


(23) 




115.5 +3.2 


138.9 + 4.9 


20.2 




(30) 


(30) 




COMBINED 


115.3 +3.3 


136.7 +5.0 


18.6 




(40) 


(53) 




EUROPEAN QUEEN X 




\ mJ,,* / 




EUROPEAN DRONE 








YD28'^ 


126.8 + 4.2 


138.7 +3.5 


9.4 


WEl'^ 


(30) 
125.1 + 4.3 


(30) 
143.3 +2.9 


14.5 


SDYl^ 


(30) 


(30) 




115.4 + 4.2 


135.0 + 6.9 


17.0 




(20) 


(19) 




COMBINED 


123.3 +6.3 


139.5 +5.4 


13.1 




(80) 


(79) 





bw-H+K J®^^®^" opposite sides of the hexagonal cell Ts 4.8 mm. 
Width between opposite sides of the hexagonal cell is 5.4 mm. 
Natural matings, multiple Inseminations. 
Single drone insemination. 



51 



TABLE 3-3. Worker bee size: hypotheses and analyses (Mann- 
Whitney U test, one-tailed, alpha = 0.05). 
Letters refer to experimental treatments, see 
Table 3-1. 



HI: Africanized bee pupae are smaller than European bee 
pupae inSependent of comb cell size. 



ACE X MOQ 


***^ 


BDF X NPR 


*** 


ACE X hPR 


*** 


BDF X MOQ 


NS 



H2: For a given egg genotype, pupae that develop in 

Africanized comb cell size are smaller than pupae 
that develop in European comb cell size. 



A X B 




*** 


C X D 




*** 


E X F 




*** 


G X H 




*** 


I X J 




*** 


K X L 




*** 


M X N 




*** 


X P 




*** 


Q X R 




*** 


ACE X 


BDF 


*** 


MOQ X 


ITR 


*** 



a«** = p<o.001. 



52 



TABLE 3-4. Reciprocal F-, cross. Pupal weights (mg), mean + SD, 

(sample size). Data are from European comb cell size only. 



AFR^ QUEEN 

X 
APR DRONE ^ 



APR QUEEN 

X 
EUR DRONE'^ 



EUR^ QUEEN 

X 
APR DRONE*= 



EUR QUEEN 

X 
EUR DRONE^ 



(A26) 



(SDA12) 



(SDYIO) 



(SDYll) 



(SDYl) 



121.7 +5.5 
(80) 



122.4 +4.4 
(40) 



133.7 +3.3 138.9 +4.9 
(23) (30) 



135.0 + 6.9 
(19) 



B 



H 



R 



^APR = Africanized; EUR = European. 

"Naturally mated. 

^Single-drone» artificial insemination. 



53 



TABLE 3-5. Maternal effect: hypotheses and analyses (Mann- 
Whitney U test, one-tailed, alpha = 0.05). 
Letters refer to experimental treatments, see 
Table 3-4. The analyses of the following 
hypotheses (a postiori) demonstrate that the pupal 
weights of hybrids from a reciprocal Ft cross are 
different from each other (H4) but are the same as 
their respective queen mothers (H2 and H6). 



HI: 


B < R 


H2: 


B < H 


H3: 


B < J 




B < L 


H4: 


H < J 




H < L 


H5: 


H < R 


H6: 


J < R 




L < R 



NS 

*** 
*** 

»#* 
*** 

*** 

NS 
NS 



^** = P<0.001. 



TABLE 3-6. Coefficients of variation for pupal weights from 
artificial, single drone inseminations and 
natural* multiple matings. 



COMB CELL SIZE 



EGG GENOTYPE^ AFRICANIZED EUROPEAN 

SINGLE DRONE INSEMINATIONS 

SDA12 3.9 3.6 

SDYIO 3.2 2.5 

SDYll 2.8 3.5 

SDYl 3.6 5.1 

YD28 3.3 2.5 

MULTIPLE INSEMINATIONS 

A26 4.4 4.5 

A57 5.8 3.4 

B39 3.1 5.1 

WEI 3.4 2.0 



analyses'^ NS 



fSee Table 3-1 for explanation of genotypes. 
°Mann-Whitney U test, one-tailed, alpha = 0.05, 



NS 



54 



CHAPTER IV 
QUEEN DEVELOPMENfT AND MATURATION 



Introduction 

African honey bees. Apt? mellifera scutellata (formerly classified 
as ^dgpsQpit; Ruttner 1976a, 1976b, 1981), were introduced into 
southeastern Brazil in 1956 (Kerr 1967; Michener 1975; Woyke 1969). The 
following year, swarms escaped and hybridized with the established 
European honey bees (primarily A. m. lioustica and mellifera ) that had 
been introduced by 1845 (Gerstaker cited in Pellet 1938; Woyke 1969). 
The descendents from this hybridization are known as Africanized honey 
bees (Goncalves 1982). 

Africanized honey bees in South America have a very high annual 
reproductive rate compared with European honey bees in temperate 
regions. Based on demographic data collected in French Guiana, the net 
reproductive rate for Africanized bees is estimated to be 16 colonies 
per colony per year (Otis 1980, 1982a). In comparison, the annual rate 
determined for European honey bees in North America was 0.92-0.96 
(Seeley 1978) or 3-3.6 when afterswarms are considered (Winston 1980a; 
Winston, Taylor and Otis 1983). This dramatic difference in 
reproduction between these two honey bee populations may be a result of 
length of time throughout the year that resources are available in the 
tropics compared with temperate regions (see Chapter VIII) and/or 



55 



56 
demographic characteristics of Africanized honey bees that account for 
high reproductive rates. 

The reproductive rate of Africanized honey bees results in a swann- 
to-swarm interval of approximately 90 days (Winston 1979b). During that 
1 period, a virgin queen emerges, develops pheromones necessary to attract 
drones, and mates; ovarian follicles mature; oviposit ion is initiated; 
and the colony population growth period begins prior to the next 
swarming. One expected demographic feature for a population with a high 
reproductive rate would be a short queen maturation interval (Fletcher 
1977a). For the Africanized queens in French Guiana, the maturation 
interval from pupal eclosion to initiation of oviposition was 9.7 days 
(Otis 1980), over 10% of their swarm-to-swarm interval (calculated from 
Winston 1979b). Fletcher and Tribe (1977b) report that in the parental 
African population, oviposition begins on the 8th to 9th day after queen 
emergence. European queens begin ovipositing between the 6th and 17th 
day after emergence (Laidlaw and Eckert 1962; Oertel 1940; Root 1947). 
Otis (1980) calculated that the mean interval from pupal eclosion to 
oviposition for European queens (10.7 days) was not significantly 
different from that of Africanized queens (9.7 days). However, 
comparisons between reported values for both Africanized and European 
honey bees are inappropriate because the data were collected under very 
different experimental conditions. Therefore, this study was undertaken 
to determine if the queen maturation interval for Africanized honey bees 
is significantly different than that for European honey bees under 
identical conditions. Three aspects of queen maturation were evaluated: 
1) larval, pupal and total development time from egg to adult emergence; 



57 
2) post-emergence development of queen attractiveness to drones; and 3) 
time from adult emergence to initiation of oviposit ion. 

In the studies reported here, queen development and maturation were 
evaluated under controlled conditions. Total development time is 
defined as the time from oviposit ion to adult emergence. These 
experimental conditions avoid the problems of previous studies that 
evaluated queen development and maturation in colonies that were 
swarming (e.g., Otis 1980). Under natural swarming conditions, queens 
are very often confined within their cells by worker bees and prevented 
from emerging for 1-10 days after pupal eel os ion (Otis 1980). 
Confinement makes calculations of development times difficult, and, 
because maturation proceeds during confinement, maturation time 
calculated from emergence to beginning of oviposition would be under 
estimated. 

Methods 
Queen Development Times 

The Africanized egg source (A26) and the Africanized cell-producing 
colonies (A37 and A43) were established from queens removed from feral 
colonies found in an area of eastern Venezuela where there were no known 
European honey bees. They were identified as Africanized honey bees by 
their behavior and characteristic comb cell size (4.5-5.0 mm wide 
between opposite sides of the hexagon, see Chapter III). The European 
egg source (Y5) and the European cell-producing colonies (19, 27 28, F, 
H and HI) came from European queens commercially produced in the 
southeastern U.S.A. and shipped to Venezuela. European colony IBR was a 
stock supplied by the U.S. Department of Agriculture Bee Breeding and 
Stock Center Laboratory, Baton Rouge, Louisiana, USA. 



58 
Eggs of known ages were collected from the Africanized (A26) and 
European (Y5) egg sources by the standard commercial queen-producing 
technique of caging the queen on an empty comb within a colony (Harp 
1973; Laldlaw 1979; see also Chapter II). After 6 hours, the combs with 
the egg samples were moved to strong Incubator-colonies for the eggs to 
develop and larvae to hatch and be fed. Very young larvae, 12-18 hours 
old, were transferred (grafted) Into beeswax queen-cell cups primed with 
royal jelly and then Introduced Into queen-cell-producing colonies 
(=nurse bee colonies) (Laldlaw 1979). All cell-producing colonies had 
large worker bee populations and were Intentionally crowded Into two 
standard Langstroth hive bodies. Queens In the cell-producing colonies 
were removed 48 hours before Introducing the grafted cells. All young, 
unsealed brood was also removed 2-4 hours before Introducing the grafted 
cells. Twenty grafted Africanized and twenty grafted European cells 
were Introduced Into each cell-producing colony. There were twenty cell 
cups to a frame, ten on the top bar and ten on the middle bar. Both 
Africanized and European larvae were grafted Into the same frame, five 
each on the top bar and five each on the middle bar. All cell cups were 
equally spaced about 8 mm apart, centered on the bars. 

In one experimental trial, the effect of nurse bee genotypes on 
queen development was evaluated by comparing queen development times for 
both Africanized and European egg genotypes In both Africanized and 
European cell-producing colonies. In another trial, development times 
for Africanized queens in Africanized and European cell-producing 
colonies were compared. In both trials, the Africanized cell-producing 
colonies had comb cell sizes characteristic of Africanized honey bees 
(4.8 mm wide; Chapter III). 



59 
The queen-eel l-producing colonies were Inspected only after the 
queen cells had been sealed in order to avoid disturbance which could 
affect development times. Once the cells are sealed, cell-producing 
colonies only maintain the appropriate temperature for the pupae to 
develop normally. On the sixth day after grafting, each sealed cell was 
protected by placing a 3 mm wire mesh tube around it to avoid any 
problems associated with queens being confined to their cells by worker 
bees. In addition, this also prevented any emerged virgin queens from 
destroying sealed cells that had not yet emerged. Beginning 24 hours 
before any expected queen emergence, the eel l-producing colonies were 
inspected daily at 0630, 1200 and 1730 hours to record queen emergence. 
In two trials, cell-producing colonies were inspected daily at 0630, 
1200 and 1730 hours, beginning 24 hours prior to estimated sealing time, 
in order to determine unsealed development times. 

Development Qf Attractiveness of Virnin Afric^ni7P d and Fi.mp^;,n Mnnpy 
Bee Queens to Drones 

Two Africanized queen mothers (A26 and A57) were removed from feral 
colonies in eastern Venezuela. The two European queen mothers (We and 
Yk) were shipped to Venezuela from different commercial queen breeders 
in southeastern USA. 

Queens from the four queen mothers (A26, AS7, We and Yk) were 
produced as described above. Sealed queen cells were removed from the 
cell-producing colony and placed in an incubator (35 + 1°C) 48 hours 
prior to emergence. After emergence, the queens were marked for 
individual identification and maintained in separate cages in a queen 
storage colony (Laidlaw 1979). 

In order to test for the degree of attractiveness to drones, each 
queen was tethered In a clean, plastic screen bag. The mesh size was 1 



60 
X 1 mm, and each bag was approximately 5 x 10 cm. Bags were 
-Individually suspended on monofilament line about 6.5 meters above the 
ground, centered between two poles 20 meters apart. Queens could be 
rapidly raised and lowered by a pulley system. Queens were put into the 
mesh bags just prior to testing in order to avoid any pheromone 
accumulation. 

The testing location was in an open field in a drone congregation 
area (Zmarlicki and Morse 1963), which was located by walking with a 
helium-filled weather balloon with mature queens suspended 10-20 meters 
above the ground. The drone congregation area was identified when 
hundreds of drones oriented to the tethered queens. Boundaries appeared 
to be quite distinct and stable through time. Both Africanized and 
European drones were probably present, but the identity of each drone 
responding to specific queens during the experiment was not known 
because there are no reliable techniques to identify individual 
Africanized and European honey bees. How drone congregation areas 
become established is not understood, but these areas are probably where 
most mating occurs. 

Individual queens were tested for drone response on consecutive 
days, beginning on the day of emergence. Only one queen at a time was 
tested so that the relative attraction of each queen would not be 
influenced by other queens being tested simultaneously. Testing lasted 
for a maximum of 3 minutes for each queen, even if no drone response was 
observed. Periodically, empty bags (blanks) were tested to insure that 
drones were responding only to the queens and not orienting to the 
experimental set-up and responding to the mesh bags. At no time did 
drones respond to the blanks. A random sequence for testing individual 



61 
queens was establ ished on each day of the experiment. Each queen was 
tested more than once on each day and always 1n a new mesh bag, to avoid 
any pheromone accumulation or contamination. Each testing session was 
begun by suspending an older queen that had previously been determined 
to be maximally attractive, in order to insure that a responding drone 
population was available. This process was also repeated if the testing 
session was interrupted by rain, extreme cloudiness, or high winds- 
conditions that normally reduce drone flight activity. The testing took 
place between 1400 and 1600 hours. 

Drone response was evaluated by assigning one of the following 
ranks to the test queen: 
Rank = no response 
Rank 1 = drones oriented to the test subject but only flew past; 

no circling of the test subject 
Rank 2 = drones oriented to the test subject and persisted in a 
wide circling formation more than 2 m from the subject 
Rank 3 = drones oriented to the test subject and formed a loose 
comet-like formation down wind more than 0.5 m to the 
test subject; formation was volatile, continually 
fragmenting and reforming; drones did not land on the 
mesh bag 
Rank 4 = drones oriented to the test subject and formed a tight 
comet-like formation down wind less than 0.5 m from the 
test subject; formation was persistent and did not 
fragment even as the test subject was lowered; drones 
landed on and walked over the mesh bag. 



62 

These ranking categories were easily discriminated and were not affected 
by the absolute numbers of drones flying. No estimates of the drone 
population were made. 
Time Post-Emergence to the In itiation of Qviposition 

Queens were produced as described above from one Africanized egg 
source (A26) and one European egg source (We). Twenty-seven Africanized 
and twenty-five European mature queen cells (two days prior to 
emergence) were each introduced into a four-frame queenless mating 
colony. Any natural queen cells in the mating colonies were destroyed 
before introducing the experimental queen cells. This insured that the 
only queen in the mating colony would be the experimental queen. When 
only one queen cell is present, worker bees usually do not confine her 
to her cell and the problem of calculating maturation time is avoided. 

Because Africanized and European queens did not develop at the same 
rate, the day of queen emergence was determined by the mean time of 
emergence for a sample of sister queens from the same graft that were 
left to emerge in an incubator at 35 + 1°C. On the eleventh day after 
the queens emerged, the colonies were inspected and the age of the brood 
was evaluated to determine the age post-emergence when the queens had 
begun ovipositing. Those colonies in which there were no larvae were 
inspected three and five days later. 

This experiment took place during the dry season. Clear weather 
prevailed so that mating flights were not affected by weather 
conditions. Both Africanized and European drones were in the area. 



63 

Res u 1 ts 
Queen D evelopment Times 

Table 4-1 presents the experimental matrix for evaluating the 
interaction of egg genotype and nurse bee genotype on queen development 
times for Africanized and European queens. Table 4-2 presents the total 
development times from oviposit ion to adult emergence for Africanized 
queens and European queens in Africanized and European cell-producing 
colonies. Table 4-3 presents the analyses for the paired comparisons in 
each cell-producing colony. These paired comparisons avoid any 
differences between cell -producing colonies because colony size (nurse 
bee population), brood area temperature, and quantity and quality of 
larval food are factors that affect queen development (Beetsma 1979; 
Laidlaw 1979; Johansson and Johansson 1973). Africanized queens develop 
in 14.5 days post-oviposit ion compared with 15.0 days for European 
queens (P<0.001, Kolmogorov-Smirnov one-tailed test, chi-square 
distribution, df = 2; Siegel 1956). There was no significant effect of 
the cell-producing colony on queen development times (Kolmogorov-Smirnov 
two-tailed test, chi-square distribution, df = 2, alpha = 0.05) (Tables 
4-4 and 4-5). 

Table 4-6 presents the development times for the Africanized queens 
in Africanized and European cell-producing colonies. There was no 
difference in Africanized queen development time between Africanized and 
European cell-producing colonies (Kolmogorov-Smirnov one-tailed test, 
chi-square distribution, df = 2, alpha = 0.05). 

The median unsealed development times from oviposition to sealing 
for both the Africanized and European queens was 7.5 days (Table 4-7). 
However, the Africanized and European genotypes were significantly 



64 
different as a result of the distribution around the median (P<0.05, 
Kolmogorov-Smirnov one-tailed test, chi-square distribution, df = 2). 
Development of Att ract ivene?;s to Dronf^ 

The response of drones to tethered, virgin queens is summarized in 
Table 4-8. There were no differences between Africanized and European 
queens with respect to either the earliest age at which a positive drone 
response (Rank 1) was observed or the earliest age at which a maximum 
drone response (Rank 4) was observed. Both Africanized and European 
virgin queens were able to attract drones (Rank 1) on the day they 
emerged. Africanized virgin queens can maximally attract drones (Rank 
4) by the fourth day post-emergence; European virgin queens can elicit a 
Rank 4 response by the fifth day post-emergence. This difference was 
not significant (Kolmogorov-Smirnov one-tailed test, chi-square 
distribution, df = 2, alpha = 0.05). The data in Table 4-8 have been 
combined for the two Africanized and two European queen lines. However, 
the queens within a population (Africanized or European) or within a 
line within a population were not uniform with respect to drone response 
or the rate of maturation. There were differences between the two 
Africanized lines and between individuals within the same line for the 
earliest age for a Rank 4 response. This same variation between lines 
and within lines existed for the European population. 
Time Post-Emergen ce to Initiation of Ovipositinn 

Table 4-9 presents the data for time post-emergence to the 
initiation of oviposit ion for both Africanized and European queens. 
Africanized queens began oviposit ion at 8.5 days post-emergence whereas 
European queens begin at 7.5 days (P<0.05, Kolmogorov-Smirnov two-tailed 
test, chi-square distribution, df = 2). 



65 
Discussion 
Queen Development 

Development ti'me from ovipos1t1on to emergence for Africanized 
queens in this study was 14.5 days, which is the same as the development 
period reported for both Africanized bees in French Guiana (Winston 
1979c) and their parental population, A. m. adansonii (= scute11ata ) in 
South Africa (Anderson, Buys and Johannsmeier no date; Fletcher 1978; 
Fletcher and Tribe 1977c). The European queens in the present study 
developed in 15.0 days, which is about one day shorter than expected 
from previous reports (Jay 1963; Laidlaw 1979). Therefore, the 
difference in development times for Africanized and European queens was 
not as great as expected and underscores the importance of making 
comparisons under the same experimental conditions. 

Queens have approximately a 25% shorter development period than 
worker bees. Differences in total development times between queens and 
worker bees are primarily due to a much shorter sealed development 
stage, i.e., 7.5 days compared with 12 days for European bees and 7.0 
days compared with 12 days for Africanized bees. The sealed development 
stage in worker bees is approximately 60% of the total development time, 
whereas in queens it is approximately 50%. 

European queens took 3.4-5.6% longer than Africanized queens to 
develop. This difference is similar to the 5.3% difference in 
development times between European and Africanized worker bees from the 
same two egg sources (A26 and Y5) (see Chapter II). 

There was no effect of nurse bee genotype on queen development 
times. However, queens were produced more successfully in European 
colonies. Africanized nurse bees were easily disturbed when the grafted 



66 
cells were Introduced into the colonies, resulting in poor survival or 
acceptance of the grafted larvae (5-50% for Africanized colonies, 
compared with 35-95% for European colonies). In addition, Africanized 
, colonies were difficult to manage because of excessive stinging that 
occurred when manipulating the strong colonies that were necessary for 
proper queen production. 

Page and Erickson (1984) found evidence that nurse bee colonies 
preferentially raised queens from more closely related larvae. However, 
in the present study, no evidence for kin recognition was observed. 
Africanized and European nurse bee colonies reared Africanized and 
European queens with equal frequency (Table 4-2). 
Rate of Matiirsti^n 

Attractiveness of queens to drones is a function of the amount of 
pheromone (9-oxodec-trans-2-enoic acid) produced in the mandibular 
glands of the queens (Butler 1971; Boch, Shearer and Young 1975). In 
England, using European genotypes, Butler (1971) tethered virgin queens 
of various ages 6 meters above the ground in areas where drones were 
flying. He determined that queens younger than 5 to 6 days old seldom 
elicited a positive drone response. Maximum positive responses from 
drones were observed in queens 8 or more days old. Butler's results 
differ from those presented in this study and may be attributed to 
either differences in experimental conditions or genetic differences 
between the queen lines studied rather than to differences due to any 
tropical or temperate conditions. The response of drones to queens 
reported in this chapter was evaluated in a drone congregation area 
which may account for the differences between the studies. 



67 

In addition. Africanized drones (not present in Butler's study) may 
have a lower response threshold to queen pheromone and therefore would 
respond to queens with less pheromone present than would European 
drones. This hypothesis is suggested by the observation that there are 
differences in the sensory receptors on the antennae of Africanized 
drones compared with European drones (Dietz 1978). Further comparison 
between Africanized and European drones is needed to determine any 
differences in the threshold of response and whether or not this would 
give the Africanized drones a mating advantage. 

Another factor that needs to be considered when using this 
behavioral bioassay (drone response) to compare rate of maturation of 
queens is that the pheromone is not continually produced but rather is 
pulsed in its production (R. Boch, pers. comm. ) . This factor may help 
to explain some of the variation of responses produced by queens within 
the same line. For example, in a few trials, a queen elicited a 
decreased response compared with the previous response she had elicited. 
In the present study, the time post-emergence to the initiation of 
oviposit ion for Africanized queens was 8.5 days and for European queens 
was 7.5 days. There are no other data available that allow for valid 
comparisons. For example, Otis (1980) reports that the mean interval 
from emergence to initiation of oviposition was 7.8 days for Africanized 
queens in French Guiana. However, these data were collected by 
observing queen maturation in colonies that had swarmed and, therefore, 
the time from emergence to oviposition would be shortened because of a 
variable period of queen confinement Cl-10 days (Otis 1980)] within the 
cells. In another set of data, Otis (1980) reports the mean maturation 
interval from eclosion to oviposition was 9.7 days. However, he does 



68 
not indicate how he determined when ecloslon occurred, or if he was 
using the terms eclosion and emergence interchangeably. The normal time 
from eclosion to emergence for queens is approximately 12 hours (Jay 
1963). 

Because of their high reproductive rate and resultant short swarm- 
to-swarm interval, Africanized honey bees were expected to have a rapid 
queen maturation interval compared with European honey bees. Fletcher 
and Tribe suggest "that in the adansonii [= scutenata 1 race, natural 
selection has worked strongly in favour of minimizing the period between 
the loss of a queen [from swarming] and the re-establishment of 
oviposit ion by a new queen" (1977b, p. 167). The surprising result from 
this study was that both Africanized and European queens matured at 
approximately the same rate, determined both by their attractiveness to 
drones and the time from adult emergence to initiation of oviposition. 
As Fletcher and Tribe (1977b) suggested, one would expect natural 
selection to be operating to minimize the maturation interval for 
queens, in order to maximize brood production between swarming periods. 
However, Africanized queens may be under a second and possibly more 
important selection pressure which may affect their maturation interval. 
Africanized swarms may travel great distances (Fletcher 1978; Michener 
1975). Otis (1980) confirmed that at least some queens issuing with 
afterswarms had already mated. If new queens issuing with these swarms 
have mated prior to swarming or mate while enroute, then delayed 
maturation, particularly with respect to development of ovarian 
follicles, would be advantageous. Follicular development would increase 
the queen's weight and make it more difficult for her to fly. Prior to 
issuing with the prime swarm, older queens usually stop egg laying 



69 

several days before the swarm departs, allowing time for their ovaries 
to recess. Therefore, maturation for Africanized queens may be delayed 
in order for the swarms with new queens to be able to migrate long 
distances. Rather than selection operating to shorten the maturation 
interval, selection may be operating to delay maturation to enable long 
swarm migration distances. 

The variation in queen maturation rates (see Tables 4-8 and 4-9) 
observed both within a population and within a queen line suggests that 
the physiological parameters involved in the process of maturation may 
be genetically determined. The rate of maturation is an important 
economic characteristic for commercial queen producers to consider in 
their selection programs. Reducing the time from emergence to 
initiation of oviposition can significantly increase the number of 
queens produced in each mating colony during the queen-producing season. 



70 



TABLE 4-1. Experimental matrix for the comparison of total 

development times (ovi'positlon to adult emergence) 
for both Africanized and European honey bee queens. 







EGG GENOTYPES 




NURSE BEE GENOTYPE^ 


AFRICANIZED 


(A26) 


EUROPEAN (Y5) 


AFRICANIZED 












A43 
A37 


A 
C 








B 
D 


COMBINED 


E 








F 


EUROPEAN 












19 

27 

28 

F 

H 

IBR 


6 
I 
K 
M 

Q 








H 
J 
L 
N 
P 
R 


COMBINED 


S 








T 


COMBINED AFRICANIZED 
AND EUROPEAN 


U 








V 



^Queen-eel 1-producing colony. 



71 



TABLE 4-2. Total development times (in days from 
oviposit ion to adult emergence) for 
Africanized and European honey bee queens: 
median, (sample size). 



EGG GENOTYPES 



NURSE BEE GENOTYPE^ AFRICANIZED (A26) EUROPEAN (Y5) 



AFRICANIZED 








A43 
A37 


14.0 
14.5 


(1) 
(7) 


15.0 (4) 
15.0 (4) 


COMBINED 


14.5 


(8) 


15.0 (8) 


EUROPEAN 








19 
27 
28 

F 
H 

IBR 


14.5 
14.5 
14.0 
14.0 
14.0 
14.0 


(15) 
(17) 
(13) 
(10) 
(7) 
(8) 


15.0 (19) 
15.0 (13) 
14.5 (12) 
14.5 (9) 
14.5 (8) 
14.8 (8) 


COMBINED 


14.2 


(70) 


15.0 (69) 


COMBINED AFRICANIZED 
AND EUROPEAN 


14.5 


(78) 


15.0 (77) 



^Queen-eel 1-producing colony. 



72 



TABLE 4-3. Analyses for the comparison of queen development 
times for both Africanized and European honey bee 
genotypes. Letters A - V represent different 
treatments! see Table 4-1 for explanation. 
Kolmogorov-Smirnov one-tailed test, chi-square 
distribution, df = 2, alpha = 0.05 (Siegel 1956). 



A 


X 


B 


NS^ 


C 


X 


D 


NS^ 


E 


X 


F 


*b 


G 


X 


H 


*** 


I 


X 


J 


** 


K 


X 


L 


** 


M 


X 


N 


** 





X 


P 


* 


Q 


X 


R 


* 


S 


X 


T 


*** 


u 


X 


V 


*** 



Small sample size, chi-square distribution is conservative. 
° * = P<0.05 
** = P<0.01 
*** = P<0.001. 



73 



TABLE 4-4. Analyses of queen development times for the 

Africanized egg genotype in the different cell- 
producing colonies. Letters represent different 
cell-producing colonies; see Table 4-1 for 
explanation. Kolmogorov-Smirnov two-tailed test, 
chi-square distribution, df = 2, alpha = 0.05 
(Siegel 1956). 



A X C NS 

A X G NS 

A X I NS 

A X K NS 

A X M NS 

A X NS 

A X Q NS 

C X G NS 

Cxi NS 

C X K NS 

C x M NS 

C X NS 

C X Q NS 

G X I NS 

G X K NS 

G X M NS 

G X NS 

G X Q NS 

I X K NS 

I X M NS 

I X NS 

I X Q NS 

K X M NS 

K X NS 

K X Q NS 

M X NS 

M X Q NS 

X Q NS 



74 



TABLE 4-5. Analyses of queen development times for the 
European egg genotype 1n the different cell- 
producing colonies. Letters represent different 
cell-producing colonies; see Table 4-1 for 
explanation. Kolmogorov-Smirnov two-tailed test, 
chi-square distribution, df = 2, alpha = 0.05 
(Siegel 1956). 



B X D 


NS 


B X H 


US 


B X J 


NS 


B X L 


NS 


B X N 


NS 


B X P 


NS 


B X R 


NS 


D X H 


NS 


D X J 


NS 


D X L 


NS 


D X N 


NS 


D X P 


NS 


D X R 


NS 


H X J 


NS 


H X L 


NS 


H X N 


NS 


H X P 


NS 


H X R 


NS 


J X L 


NS 


J X N 


NS 


J X P 


NS 


J X R 


NS 


L X N 


NS 


L X P 


NS 


L X R 


NS 


N X P 


NS 


N X R 


NS 


P X R 


NS 



75 



TABLE 4-6. Total development time (in days from oviposltion 
to adult emergence) of Africanized queens in 
Africanized and European cell-producing colonies: 
median, (sample size) . 



nurse bee genotype^ africanized egg genotype (a26) 
africanized"^ 

A43 14.4 (10) A 

A37 14.6 (6) B 

A43 & A37 14.4 (16) C 
EUROPEAN 

HI 14.4 (16) D 

IBR 14.2 (14) E 

HI & IBR 14.4 (30) F 



ANALYSES^ 



A X D NS 

A X E NS 

B X D NS 

B X E NS 

C X F NS 



^Queen-cell-producing colonies. 
"Africanized comb cell size. 

^Kolmogorov-Smirnov one-tailed test» chi-square 
distribution, df = 2, alpha = 0.05. 



76 

TABLE 4-7. Unsealed (egg and larval periods combined) development times 
(in days) for Africanized and European queens: median, 
(sample size) . 



NURSE BEE GENOTYPE^ 



19 
28 

19 & 28 



EGG GENOTYPES 



EFFECT OF EGG 
AFRICANIZED (A26) EUROPEAN (Y5) GENOTYPE^ 



7.5 (17) 
7.2 (14) 
7.5 (31) 



7.5 (19) 
7.5 (12) 
7.5 (31) 



EFFECT OF NURSE BEE 
GENOTYPE^ 



NS 



NS 



Cell-producing colonies, European nurse bees, European comb cell size. 
Kolmogorov-Smirnov one-tailed test, chi-square distribution, df = 2, 
* = P<0.05. 
^Kolmogorov-Smirnov two-tailed test, chi-square distribution, df = 2, 
alpha = 0.05. 



77 



TABLE 4-8. Drone response to tethered virgin queens: median 
day of response post-emergence, range, (sample 
size). 



DAY OF RESPONSE LEVEL 





RAI^ 1 


RANK 2 


RANK 3 


RANK 4 


AFRICANIZED QUEEN 
GENOTYPES 

EUROPEAN QUEEN 
GENOTYPES 


0^ 
(2) 

0^ 
(1) 


3.5 
1-5 
(6) 

1.5 
1-5 
(4) 


3.5 

1-5 
(6) 

4.0 
2-5 
(3) 


4.0 
1-5 
(5) 

4.5 
4-5 
(2) 


analyses'' 


— 


NS 


NS 


NS 



^Day = day of adult emergence. 

Kolmogorov-Smirnov one-tailed test, chi-square distribution, 
df = 2, alpha = 0.05. 



78 



TABLE 4-9. Time post- emergence to initiation of oviposition: 
median, range, (sample size) . 



DAYS POST-EMERGENCE 



AFRICANIZED GENOTYPE (A26) 8.5 

7.5-12 
(10) 



EUROPEAN GENOTYPE (We) 7.5 

6-10 
(16) 



ANALYSIS^ P<0.05 



Kolmogorov-Smirnov two-tailed test, chi-square distribution, 
df = 2. 



CHAPTER V 
QUEEN PUPAL WEIGHTS 

Introduct-fon 
Africanized honey bees In South America are hybridized descendents 
of African honey bees (A pis melUfera scutellata ) and European honey 
bees (primarily A. on.. 11gu?tica and A. in. melllfera ) (Goncalves 1982; 
Woyke 1969). The annual net reproductive rate of Africanized honey bees 
1n South America Is four to five times greater than that of European 
honey bees In temperate regions: 16 colonies per colony per year 
compared with 3-3.6 (Otis 1980, 1982a; Winston 1980a; Winston, Taylor 
and Otis 1983). Differences In reproductive rates between these two 
honey bee populations may be a result of: 1) colony demography; 
2) temperate vs. tropical climate and floral resources; 3) resource 
utilization behaviors; or 4) a combination of factors. Because 
Africanized and European honey bees have not been compared under 
Identical experimental conditions. It Is not possible to determine to 
what extent reproductive differences are a result of genetic or 
environmental parameters. 

One demographic parameter associated with rapid colony growth and a 
high rate of colony reproduction would be a high oviposltlon rate (Brian 
1965; Moeller 1961; Wilson 1971). In the evolution of social Insects, 
queen oviposltlon rates have increased primarily due to one of the 
following: increased number of ovarioles, increased length of the 

79 



80 
ovan'oles, more rapid egg maturation, and reduction in egg size (Hagan 
1954 and Iwata and Sakagami 1966 cited in Wilson 1971; Wilson 1971). 
Honey bee queens have a very large number of ovarioles (>300) and, for 
European queens, the number of ovarioles has been shown to be an 
inherited character (Eckert 1934) which is positively correlated with 
queen pupal weight (Hoopingarner and Farrar 1959). Queen weight was 
also found to be correlated with brood production (Boch and Jamieson 
1960). If it is assumed that both Africanized and European honey bees 
have the same relationship between queen weight and brood production, 
then weights of queens from the two populations can be compared to 
determine potential differences in fecundity. 

In honey bees, differentiation between worker and queen castes is 
not genetically determined, but rather is regulated by the quantity and 
quality of food fed to developing larvae during the first 3 days 
(Beetsma 1979). Therefore, a number of factors other than genotype 
affect queen size, e.g., age of larvae used to produce queens, 
population of the cell-producing colony, quantity and quality of food 
fed to developing larvae, and temperature (Beetsma 1979; Johansson and 
Johansson 1973; Laidlaw 1979; Weiss 1974; Woyke 1971). Because of 
differences in queen rearing methods and experimental conditions, 
previous comparisons of size between Africanized and European queens may 
be inappropriate. This study was undertaken to compare queen pupal 
weights for Africanized and European honey bees under identical 
experimental conditions in Venezuela. 

Methods 
Four Africanized honey bee lines (A26, A57, A61 and A62) were 
established from queens removed from feral colonies in an area in 



81 
eastern Venezuela that had no known European honey bees. They were 
Identified as Africanized honey bees by their comb cell size, which was 
significantly smaller than European comb cell size (Chapter III). Two 
European lines (YK and WE) were established from queens shipped to 
Venezuela by commericial queen producers in the southeastern U.S.A. 
Three additional European lines (YD, N and GK) were established from 
queens shipped to Venezuela from the U.S. Department of Agriculture Bee 
Breeding and Stock Center Laboratory, Baton Rouge, Louisiana, U.S.A. 

Queens were produced from these nine lines by standard queen 
rearing methods (Laidlaw 1979). Egg samples from the nine queen mothers 
were collected by confining the queens to an empty comb within their own 
colonies using an 8 x 8 cm push- in cage made from 3 mm mesh hardware 
cloth. -Queen excluder material was soldered to the tops of the push-in 
cages, allowing worker bees to move in and out in order to feed and tend 
the queen (Harbo, Bolten, Rinderer and Collins 1981). Both Africanized 
and European eggs were collected in European size comb. After 
approximately 4-6 hours, the queens were released, and combs containing 
the eggs were put into a strong colony in order for the eggs to be 
incubated and for the larvae to be fed. Africanized and European eggs 
were both put into the same incubator-colony in order to control for any 
differences in early larval feeding and temperature. 

Young larvae approximately 12-15 hours old were transferred 
(grafted) into artificial, beeswax, queen-cell cups and then introduced 
into the cell-producing colonies. Twenty larvae from one of the 
Africanized lines and twenty larvae from one of the European lines were 
grafted into each cell -producing colony. To control for extrinsic 
factors affecting queen size, analyses of Africanized and European 



82 

queens were limited to paired comparisons (one Africanized and one 
European line) that were each simultaneously introduced into the same 
incubator-colony and then grafted into the same queen-cell-producing 
colony. Possible effects from different cell-producing colonies on 
queen pupal weight were evaluated by grafting the same queen lines into 
different cell-producing colonies. 

Only European cell-producing colonies were used because of the 
difficulty in producing queens in Africanized colonies. Africanized 
cell-producing colonies remained disturbed for a long period of time 
after the grafted larvae were introduced, which resulted in poor 
acceptance (survival) of the larvae (see Chapter IV). 

Queen pupal weights are used for comparison because adult weights 
vary with respect to engorgment of food, dehydration, feces 
accumulation, and differential ovariole development. Although queen 
pupal weights vary with age of the pupae, there is a period from the 
10th through the 13th day post-oviposit ion when queen pupal weight is 
constant (Table C-1). Queen pupal weight comparisons can therefore be 
made during this period (Hoopingarner and Farrar 1959). Although there 
is a 0.5 day difference in development time between Africanized and 
European queens (Chapter IV), the 3-4 day pupal period during which 
there is no significant weight change is of sufficient duration to allow 
Africanized and European queens to be accurately and consistently 
compared. Africanized and European queen pupae were weighed on the 11th 
day post-oviposit ion. Queen cells from each of the lines were randomly 
selected to avoid any position effect from location on the grafting 
frame. Weights were measured to the nearest 1.0 mg using either a 
Mettler Type H4 or H6 balance. 



83 

Queen cell lengths were measured at the time the queen pupal 
weights were determined. A calipers was used to determine the external 
length from the base to the apex of the queen cell. 

Results ' 
Queen pupal weights for four Africanized and five European lines 
are presented in Table 5-1. European queen pupal weights were 
significantly larger than Africanized queen pupal weights for two 
different pairwise comparisons, (YK vs. A26 and YK vs. A57; P<0.05 to 
P<0.001j Mann-Whitney U test, one-tailed). Africanized queen pupal 
weights were significantly larger in one pairwise comparison (A62 vs. N; 
P<0.02; Mann-Whitney U test, two-tailed). For three pairwise 
comparisons, there was no statistical difference (A26 vs. WE, A57 vs. 
YD, and A61 vs. GK; Mann-Whitney U test, one-tailed, alpha = 0.05). 
Because different cell-producing colonies had no significant effect on 
queen pupal weights (see below), the means for the nine queen lines can 
be ranked and analyzed (Table 5-2). There was no significant difference 
between the Africanized and European honey bee populations for queen 
pupal size (Mann-Whitney U test, one-tailed, alpha = 0.05). 

Queen cell lengths for Africanized and European lines are presented 
in Table 5-3. In six out of eight pairwise comparisons, there was no 
significant difference in queen cell lengths between Africanized and 
European queens (Mann-Whitney U test, one-tailed, alpha = 0.05). For 
the pair in cell-producing colony 2, the European line was significantly 
larger than the Africanized line (P<0.05). For the pair in cell- 
producing colony 4, the Africanized line was significantly larger than 
the European line (P<0.02; Mann-Whitney U test, two-tailed). 



84 

Spearman's rank correlation coefficient was determined for queen 
pupal weights and queen cell lengths (Table 5-4). In general, there was 
no significant correlation between queen pupal weight and queen cell 
length (alpha = 0.05). However, one Africanized line (A26) In cell- 
producing colony 2 had a significant correlation (P<0.05) and one 
European line (YK) In cell-producing colony 4 had a significant 
correlation (P<0.01). 

The effect of cell-producing colonies on queen pupal weights and 
queen cell lengths Is presented In Table 5-5. There was no significant 
difference for Africanized queen line A26 in four different cell- 
producing colonies (one-way analysis of variance, alpha = 0.05); nor was 
there a significant difference for Africanized queen line A57 in two 
different cell-producing colonies (Mann-Whitney U test, two-tailed, 
alpha = 0.05). There was no significant difference in pupal weights for 
the European queen line YK in four different cell-producing colonies 
(one-way analysis of variance, alpha = 0.05), but there was a 
significant effect of cell-producing colonies on queen cell length 
(P<0.001). When cell-producing colony 2 was removed from the analysis, 
there was no significant difference in queen cell length. 

Discussion 
If we assume for both Africanized and European honey bees that 
queen weight is correlated with egg production or fecundity (Boch and 
Jamieson 1960), we would then expect that egg laying rates would follow 
the same ranking as presented in Table 5-2 for queen pupal weights. 
Based on these pupal weights, we would predict that there would be no 
difference in egg laying rates for the Africanized and European honey 
bee populations. In fact, when egg laying rates for Africanized and 



85 
European honey bee queens were compared, there was no significant 
difference between queens from the two populations (Chapter VI). There 
were, however, significant differences in pupal weights between 
individual queen lines both between and within each population (Table 5- 
1). There were also significant differences in egg laying rates between 
individual queen lines both between and within the two populations 
(Chapter VI). 

In this study. Africanized queens were reared in European colonies 
because of low acceptance (survival) of grafted cells in Africanized 
colonies (Chapter IV). Because queen-worker caste differentiation in 
honey bees is regulated by larval feeding (Beetsma 1979), rearing 
Africanized queens in European colonies may have obscured differences in 
pupal weights between Africanized and European queens. Possibly, 
European worker bees may rear larger Africanized queens than would 
Africanized worker bees because European worker bees are themselves 
larger (Chapter III), and may feed the developing queen larvae 
differently. Although virgin European queens have been reported to 
weigh more than virgin Africanized queens— 208 vs. 199 mg (Goncalves, 
Kerr and Nocoes 1972 cited in Michener 1975)— there was no indication of 
conditions under which the queens were reared. 

Further analysis of queen weights between Africanized and European 
honey bees is needed, preferably in a 2 x 2 experimental design: 
Africanized and European queens reared in both Africanized and European 
cell-producing colonies. In addition, the relationship between queen 
pupal weights and brood production needs to be evaluated for both 
Africanized and European honey bee lines to determine if the same 
relationship exists for both populations. 



86 

The European queen lines evaluated were a diverse representation of 
the European population from North America, whereas the Africanized 
queen lines may only reflect a small portion of the Africanized 
population. The location for the sources of the Africanized lines was 
limited to feral colonies found in one area of eastern Venezuela. A 
greater diversity of Africanized lines needs to be evaluated in order to 
be able to generalize about queen pupal weights and oviposition rates 
for the population as a whole. 



87 



TABLE 5-1. Comparison of Africanized and European queen pupal weights 
(mg): mean + SD, (sample size), (genotype). 

AFRICANIZED EUROPEAN 

CELL BUILDER^ GENOTYPES GENOTYPES ANALYSES^^ 



257.1 t 7.5 286.4 + 12.0 

(9) (A26) (9) (YK) 

255.8 + 9.2 284.4 +20.6 

(9) (A26) (9) (YK) 



8 237.2 +21.7 232.9 + 0.4 

(5) (A61) (2) (GK) 



*** 



»« 



3 262.3 + 6.4 266.7 + 9.7 MS 

(3) (A26) (9) (WE) 

4 260.0 + 9.0 272.5 + 16.5 « 

(11) (A57) (13) (YK) 

5 243.0 +13.4 282.6 + 12.0 * 

(3) (A26) (3) (YK) 

6 248.6+14.3 264.0+15.4 NS 

(4) (A57) (5) (YD) 

7 291.8 + 7.9 257.3 + 18.2 — <= 

(5) (A62) (4) (N) 



NS 



Cell-producing colonies; European nurse bees and European comb 
cell size. 

Mann-Whitney U test, one-tailed, alpha = 0.05' 
^* = P<0.05, ** = P<0.01, *** = P<0.001. 

Difference In wrong direction for one-tailed test; two-tailed 
test results In a P<0.02. 



88 



TABLE 5-2. Queen pupal weights for the nine lines 
analyzed. 



POPULATION 


QUEEN LINE 


MEAN PUPAL WEIGHT(MG) 


European 


GK 


233 


Africanized 


A61 


237 


Africanized 


A26 


256 


European 


N 


257 


Africanized 


A57 


257 


European 


YD 


264 


European 


WE 


267 


European 


YK 


280 


Africanized 


A62 


292 


ANALYSIS^ 




NS 



^ann-Whltney U test, one-tailed, alpha = 0.05. 



89 



TABLE 5-3. Comparison of Africanized and European queen cell 
lengths (mm): mean + SD, (sample size)» 
(genotypes) . 



CELL BUILDER^ 


AFRICANIZED 
GENOTYPES 


EUROPEAN 
GENOTYPES 


analyses'^ 


1 


2.58 
(9) 


+ 0.1 
(A26) 


2.50 
(9) 


+ 0.1 
(YK) 


NS 


2 


2.62 
(9) 


+ 0.2 
(A26) 


2.70 
(9) 


+ 0.1 
(YK) 


* 


3 


2.54 
(3) 


+ 0.1 
(A26) 


2.48 
(9) 


+ 0.1 
(WE) 


NS 


4 


2.58 
(7) 


+ 0.1 
(A57) 


2.44 
(11) 


+ 0.1 
(YK) 


c 


5 


2.67 
(3) 


+ 0.1 
(A26) 


2.57 
(3) 


+ 0.1 
(YK) 


NS 


6 


2.50 
(4) 


+ 0.02 
(A57) 


2.57 
(5) 


+ 0.1 
(YD) 


NS 


7 


2.78 
(5) 


+ 0.04 
(A62) 


2.81 
(3) 


+ 0.1 
(N) 


NS 


8 


2.79 
(5) 


+ 0.1 
(A61) 


2.76 
(2) 


+ 0.1 
(GK) 


NS 


^Cell-producing 
cell si^e. 


colonies; European 


nurse bees and 


European comb 



— _ 

■^Mann-Whitney U test, one-tailed, alpha = 0.05; * = P<0.05. 
'^Difference in wrong direction for one-tailed test; two-tailed 
test results in a P<0.02. 



90 



TABLE 5-4. Correlation of queen cell length (mm) and queen pupal 
weight (mg): mean + SD, (sample size) . Measurement 
made on day 11.25 post-oviposition. 



QUEEN CELL QUEEN PUPAL 

QUEEN GENOTYPE LENGTH WEIGHT CORRELATIONS^ 



AFRICANIZED 








A26 


(CBl)'^ 


2.58 +0.1 
(9) 


257.1 + 7.6 
(9) 


NS 


A26 


(CB2) 


2.62 +0.2 
(9) 


255.8 + 9.2 
(9) 


* 


A26 


(CBS) 


2.54 +0.1 
(3) 


262.3 + 6.4 
(S) 


— 


A26 


(CBS) 


2.67 + 0.1 
(S) 


243.0 + 13.4 
(3) 


~ 


A57 


(CB4) 


2.58 + 0.1 
(7) 


259.1 + 10.8 
(7) 


NS 


A57 


(CB6) 


2.50 +0.02 
(4) 


248.6 + 14.3 
(4) 


NS 


A62 


(CB7) 


2.78 + 0.04 
(5) 


291.8 + 7.8 
(5) 


NS 


A61 


(CB8) 


2.79 +0.1 
(5) 


237.2 + 21.7 
(5) 


NS 


EUROPEAN 








YK 


(CBl) 


2. SO + 0.1 
(9) 


286.4 + 12.0 
(9) 


NS 


YK 


(CB2) 


2.70 + 0.1 
(9) 


284.4 + 20.6 
(9) 


NS 


YK 


(CB4) 


2.44 +0.1 
(11) 


272.2 + 17.8 
(11) 


** 


YK 


(CBS) 


2.57 +0.1 
(3) 


282.6 + 12.0 
(3) 


~ 


WE 


(CBS) 


2.48 + 0.1 
(8) 


268.4 + 8.8 
(8) 


NS 


YD 


(CB6) 


2.57 + 0.1 
(5) 


264.0 + 15.4 
(5) 


NS 


N 


(CB7) 


2.81 +0.1 
(3) 


264.3 + 14.0 
(3) 


— 


GK 


(CBS) 


2.76 +0.1 
(2) 


232.9 + 0.4 
(2) 


^^ 



^Spearman's rank correlation coefficients, alpha = 0.05; 
* = P<0.05; ** = P<0.01. 

CB = cell-producing colony number; refer to Table 5-1 for 
explanation. 



91 



TABLE 5-5. 



Effect of cell-producing colony on queen cell length and 
queen pupal weight: mean + SD, (sample size). 



AFRICANIZED QUEEN GENOTYPE (A26) 

CBl^ CB2 



CB3 



CB5 analyses'^ 



QUEEN CELL 
LENGTH 


2.58 
+ 0.1 
(9) 


2.62 
+ 0.2 
(9) 


2.54 
+ 0.1 
(3) 


2.67 
+ 0.1 
(3) 


NS 


QUEEN PUPAL 
WEIGHT 


257.1 

+ 7.6 

(9) 


255.8 

+ 9.2 

(9) 


262.3 

+ 6.4 

(3) 


243.0 
+ 13.4 
(3) 


NS 



AFRICANIZED QUEEN GENOTYPE (A57) 





CB4 


CB6 


ANALYSES*^ 


QUEEN CELL 


2.58 


2.50 


NS 


LENGTH 


+ 0.1 


+ 0.02 






(7) 


(4) 




QUEEN PUPAL 


260.0 


248.6 


NS 


WEIGHT 


+ 9.0 


+ 14.3 






(11) 


(4) 





EUROPEAN QUEEN GENOTYPE (YK) 

CBl CB2 



CB4 



CB5 analyses'^ 



QUEEN CELL 


2.50 


2.70 


2.44 


2.57 


LENGTH 


+ 0.1 


+ 0.1 


+ 0.1 


+ 0.1 




(9) 


(9) 


(11) 


(3) 


QUEEN PUPAL 


286.4 


284.4 


272.5 


282.6 


WEIGHT 


+ 12.0 


+ 20.6 


+ 16.5 


+ 12.0 




(9) 


(9) 


(13) 


(3) 



***'■ 



NS 



CB - cell-producing colony number; refer to Table 5-1 for explanation. 
One-way analysis of variance, alpha = 0.05; *** = P<0.001. 
Mann-Whitney U test, two-tailed, alpha = 0.05. 
°With CB2 removed, ANOVA is NS. 



CHAPTER VI 
EGG LAYING AhO BROOD PRODUCTION RATES DURING THE FIRST BROOD CYCLE 

Introduction 
Africam'zed honey bees 1n South America are descendents from the 
hybridization of African honey bees (Aajs mellifera scutellata ) and 
European honey bees (primarily A. m.. linustir^ and A. m. meTlifer. ) 
(Goncalves 1982; Woyke 1969). In tropical and sub-tropical regions of 
South America, Africanized bees have been more successful than European 
bees as determined by their rapid rates of dispersal and high population 
densities (Michener 1975; Taylor 1977, 1985). It is not surprising that 
Africanized bees are more successful in these regions because they are 
descendents of honey bees that evolved under similar tropical conditions 
1n Africa. 

Rates of dispersal and population densities achieved by Africanized 
honey bees require a high colony reproductive rate. Africanized honey 
bees in South America have a reproductive rate that is four to five 
times greater than the reproductive rate of European honey bees in North 
America (Otis 1980, 1982a; Winston 1980a; Winston, Taylor and Otis 
1983). However, based on this comparison, one cannot identify the 
factors that account for differences in reproductive rates nor identify 
the factors leading to the success of Africanized bees in South America. 
Because the comparison of reproductive rates was not based on data 
collected under similar environmental or experimental conditions, it is 

92 



93 

not possible to determine to what extent reproductive differences 
between the two populations are a result of differences in intrinsic 
demographic parameters and/or environmental differences due to temperate 
vs. tropical resources and climatic conditions. In addition, 
experimental conditions were very different. For example, an important 
variable affecting reproductive rates in honey bees is brood-nest 
crowding (Baird and Seeley 1983; Simpson 1966, 1973; Simpson and Riedel 
1963). Experimental Africanized colonies in South America were 
maintained in 22-liter hives (Otis 1980; Winston 1979b), whereas, 
experimental European colonies in North America, with which they were 
compared, were maintained in 42-liter hives (Winston 1980a). 

Reproductive rates in honey bees are a result of an interaction of 
at least three factors: colony demography, resource availability, and 
resource utilization efficiency. As part of a larger investigation 
comparing intrinsic demographic factors between Africanized and European 
honey bees to determine which aspects of demography, if any, are 
responsible for the success of Africanized honey bees in South America, 
this study evaluated one aspect of demography—queen fecundity. Queen 
egg laying rate is one of the primary demographic parameters that 
affects colony growth rates (Brian 1965; Moeller 1961; Wilson 1971). 
Although differences in egg laying rates between Africanized queens and 
European queens have been reported (Fletcher 1978; Michener 1972, 1975; 
Ribbands 1953), they cannot be compared because the data were collected 
under different resource and experimental conditions. Therefore, the 
present study was undertaken to compare egg laying and brood production 
rates for both Africanized and European queens under identical, tropical 
conditions in Venezuela. The experimental design allowed for a 



94 

comparison between the two honey bee populations during the first brood 
cycle. Differences in initial colony growth rates between Africanized 
and European honey bees may be an important factor in determining 
differences in reproductive rates. 

Egg laying and brood production rates for Africanized and European 
honey bees were evaluated at both the queen and worker bee levels. The 
interactions of both Africanized and European queens with both 
Africanized and European worker bees were evaluated because of potential 
behavioral and/or physiological differences between Africanized and 
European nurse bees with respect to affecting brood production rates 
and/or the queen's oviposition behavior. In order to compare egg laying 
and brood production rates between Africanized and European honey bees, 
four variables needed to be controlled. 

First is colony size, because egg laying rates are positively 
correlated with the number of worker bees in a colony (Moeller 1958). 
In order to evaluate initial colony growth, a colony size was selected 
that contained the number of worker bees within the range reported for 
both Africanized and European swarms (Fell et. al. 1977; Otis 1980; 
Rinderer, Collins, Bolten and Harbo 1981; Rinderer, Tucker and Collins 
1982; Winston 1980a). 

Second is hive cavity volume, which must be controlled in order to 
avoid effects of differential brood-nest crowding on oviposition rates 
(Brian 1965). A hive cavity volume was selected that represents natural 
nest cavity volumes chosen by these two honey bee populations (Rinderer, 
Collins, Bolten and Harbo 1981; Rinderer, Tucker and Collins 1982; 
Seeley 1977; Seeley and Morse 1976; Winston, Taylor and Otis 1983). 



95 
A third variable, comb cell size, had to be controlled. If bees 
were allowed to build their own comb, comb built by Africanized workers 
would be smaller than comb built by European bees (Chapter III). The 
larger, European comb was selected for these experiments for two 
reasons: 1) European queens do not lay eggs in a uniform pattern in 
Africanized comb; and 2) there is a higher brood mortality in colonies 
with European nurse bees on Africanized comb — possibly because the 
larger European nurse bees have difficulty feeding the developing larvae 
in the smaller. Africanized cells (Chapters II and III). On the other 
hand. Africanized queens and worker bees appear to behave normally when 
managed on European comb. Only Africanized bees that had been reared in 
managed colonies with European combs were used in the experimental 
colonies to avoid any delay in adjusting to larger comb cell size. 

Finally, the fourth variable controlled was resource availability. 
Africanized and European queens were compared simultaneously so that 
floral resource conditions were identical. Surplus honey was also 
provided for each experimental colony to reduce the effects of 
differential foraging success between Africanized and European honey 
bees in tropical resource conditions (Rinderer, Bolten, Collins and 
Harbo 1984; Rinderer, Collins and Tucker 1985). 

In addition to colony-level variables, conditions under which 
experimental queens are produced may affect queen fecundity, e.g., age 
of larvae used to produce queens, quantity and quality of food fed to 
developing queen larvae, and temperature during development (Beetsma 
1979; Laidlaw 1979; Weiss 1974; Woyke 1971). The age of a queen may 
also affect her fecundity (Ribbands 1953). In this study, all 



96 
experimental queens were produced under identical conditions and were 
the same age. 

Estimates of daily egg laying rates derived from total brood 
production may not be accurate (Merrill 1924) because mortality of 
unsealed brood (eggs and larvae) may be quite high, up to 50% (Garofalo 
1977; Merrill 1924; Woyke 1977), particularly during the initial period 
of colony growth (Winston, Dropkin and Taylor 1981). Therefore, both 
daily egg laying rates and brood production during the first brood cycle 
were analyzed. 

Method? 
Daily Eaa Laving Rat.eg 

Three different Africanized queen mothers (A57, A61 and A62) were 
established from feral colonies that were found in an area of eastern 
Venezuela with no known European honey bees. They were confirmed as 
Africanized honey bees by their behavior and comb cell size (Chapter 
III). Three different European queen mothers were shipped to Venezuela 
from the U.S.A.; two (GK and YD) were from the U.S. Department of 
Agriculture Bee Breeding and Stock Center Laboratory in Baton Rouge, 
Louisiana, and the third (YK) was from a commercial queen producer in 
southeastern U.S.A. 

Experimental queens were produced from the queen mothers using 
standard queen rearing techniques (Laidlaw 1979). Eggs from the queen 
mothers were collected by restricting queens to a portion of comb for 4 
to 6 hours under 8 x 8 cm push-in cages that had queen excluder material 
soldered to the tops (Harbo, Bolten, Rinderer and Collins 1981). Combs 
with the egg samples were then placed into a populous colony where the 
eggs were incubated until the larvae hatched and were fed. Young 



97 
larvae, approximately 15 hours old, were transferred (grafted) Into 
beeswax queen-cell cups that had been primed with diluted royal jelly 
and then introduced into cell -producing colonies for development. Three 
days prior to adult emergence, sealed queen cells were put into an 
incubator (35 + 1° C). Emerged, virgin queens were marked for 
individual identification, using color-coded, plastic, numbered discs 
glued to the queen's thorax (Smith 1972). Virgin queens were maintained 
separately in small, two-frame colonies during the period of maturation 
and after artificial insemination in order to maximize the number of 
spermatozoa that migrate to the spermatheca (Woyke 1979). 

Queens were artificially inseminated one week after emergence using 
the apparatus designed by Harbo (1979) and Mackensen (Mackensen and 
Roberts 1948; Mackensen and Tucker 1970). Queens were inseminated with 
2.5 ul of wild-type semen on each of two occasions, 3 days apart, to 
increase the percentage of spermatozoa entering the spermatheca (Bolten 
and Harbo 1982; Mackensen 1964). 

Four colony treatments were established: 

1. Africanized queen with Africanized worker bees, 

2. Africanized queen with European worker bees, 

3. European queen with Africanized worker bees and 

4. European queen with European worker bees. 

Each experimental colony was in a five-frame hive (23 liters) that 
contained three empty combs of European cell size. A different 
geometric design was painted over the entrance of each hive in order to 
aid in orientation and reduce drifting of foragers between colonies (von 
Frisch 1967). 



98 
Two days prior to the beginning of the experiment, young bees were 
removed from brood frames of large colonies and put into screened cages 
that measured 48 x 37 x 76 cm. One cage contained Africanized bees and 
the other contained European bees. Africanized bees were removed only 
from colonies that were being managed on combs of European cell size. 

The cages were supplied with feeders containing 50% (volume:volume) 
sugar syrup. Bees used to stock the experimental colonies were taken 
from the appropriate cage, thereby insuring that all experimental 

colonies were uniform in composition for a particular worker bee type. 

Africanized or European. Each experimental colony was started with 

approximately 775 grams of bees. 

Test queens were introduced into each experimental colony using a 

push-in cage (Laidlaw 1979). Queens were manually released from the 

push- in cages after two days. 

The experiment consisted of two separate trials. Trial 1 evaluated 

colony treatments 2 and 4 only. Trial 2 evaluated all four colony 

treatments simultaneously. Trial 2 was started with a new supply of 

worker bees. Some of the queens used in trial 2 had also been evaluated 

in trial 1. 

Egg laying rates were determined by removing the frames from each 
colony every 24 hours and counting the number of eggs. The removed 
frames were immediately replaced with empty frames in order to minimize 
disturbance. The frames with eggs were stored in a freezer until the 
eggs were counted. Initial egg laying rates that were more than two 
standard deviations from the mean were not used because these rates may 
have occurred before queen maturation was complete. Experimental 
colonies with Africanized worker bees (trial 2) began absconding 



99 
(abandoning the hive) after five to six days because of the disturbance 
caused by the experimental procedure. The experiment was terminated at 
that point and only egg laying rates prior to the beginning of 
absconding were compared. 
Brood Production Rates diirinr^ First Rrnnd r.yr- \^ 

The same queens evaluated in the egg laying experiment were used 
for this experiment. They were maintained in separate cages in a queen 
storage colony (Laidlaw 1979; Reid 1975) for six days between 
experiments. Four colony treatments were established as described 
above, except that the hives were larger and were stocked with more 
worker bees. Standard Langstroth hive bodies (48 liters) were used with 
nine frames (all with European comb cell size): eight empty, drawn 
combs and one filled with honey. All frames were weighed prior to being 
put into the colonies in order that weight changes could be monitored. 
Different geometric designs were placed above the entrances to 
facilitate orientation and reduce drift between colonies. 

As described for the first experiment, young worker bees were 
collected and maintained in screened cages for two days prior to the 
beginning of the experiment. Approximately 1200 grams of worker bees 
were used to stock each colony. The number of bees put into each colony 
was estimated by determining the mean individual bee weight from three 
20-30 bee samples for each cage and then dividing the total weight of 
introduced bees by the mean individual bee weight (Otis 1982b). 

Queens were introduced into the experimental colonies using push-in 
cages and manually released two days later in order to standardize the 
starting day (day 0) for all experimental colonies. Queens that had 
been in Africanized colonies for the daily egg laying rate experiment 



100 
were now introduced into European colonies; queens that had been in 
European colonies during the first experiment were now introduced into 
Africanized colonies. 

After day 12 of oviposition, each experimental colony was inspected 
and the amounts of unsealed brood (eggs and larvae) and sealed brood 
(pre-pupae and pupae) for each colony were determined. To facilitate 
measuring the amounts of unsealed and sealed brood, a 2.5 x 2.5 cm grid 
was placed over each frame and the amount of brood within each square 
was estimated. The number of developing brood cells was determined by 
multiplying the brood area (in cm^) by 4.25 (the number of comb cells 
per cm ). The number of worker bees present at day 12 was estimated by 
assuming a constant rate of mortality of adults from day to day 17. 

At the end of day 17 of oviposition, colonies were closed and 
killed with potassium cyanide. Unsealed and sealed brood, numbers of 
adult bees, amount of pollen, and frame weights were determined. 
The number of adult bees present at day 17 was estimated by 
determining the mean individual bee weight from three samples of 150-200 
bees taken from each colony. The total weight of bees in each colony 
was then divided by the mean individual weight to get an estimate of the 
total number of bees in each colony (Otis 1982b). The accuracy of this 
technique was determined by comparing the estimates with the actual 
counts (Table D-1). A mean difference of only 1.5% was observed. 

Data were analyzed using the Mann-Whitney U test (alpha = 0.05). 
Correlations were evaluated using Spearman rank correlation coefficients 
(alpha = 0.05). 



101 
Results 
Daily Ena Lavfnn R^tP^ 

There were no significant differences in daily egg laying rates 
between Africanized and European queens (Tables 6-1 to 5-5). There 
were, however, significant differences between individual queen lines 
both between the two populations and within each population (Table 6-6). 
There was no significant effect of worker bee type (Africanized vs. 
European) on egg laying rates for either Africanized or European queens 
(Table 6-5). Worker bee type also had no effect on daily egg laying 
rates of sister queens (Table 6-7). 
Brood Production Rates during First Rrnnd C.yr]^ 

The amounts of unsealed, sealed and total brood at day 12 and at 
day 17 for each of the experimental colonies are presented in Table 6-8. 
There was no significant effect of worker bee population type on brood 
production at either day 12 or day 17. When all colony treatments were 
combined, total brood produced at day 12 was significantly correlated 
with total brood produced at day 17 (P<0.05). 

Changes in worker bee population for each colony are shown in Table 
6-9. The estimated daily mortality rates for Africanized worker bees 
were not significantly different than those for European worker bees. 

The numbers of unsealed brood, sealed brood and total brood at day 
12 and day 17 expressed as percentages of the adult population are 
presented in Table 6-10. There was no significant difference between 
Africanized and European worker bees. 

Sister queens are compared in Table 6-11. The performance of the 
European queen pair was similar with either Africanized and European 
worker bees. One sister of Africanized queen pair (A62) performed 



102 
better with Africanized bees, and one sister of Africanized queen pair 
(A57) performed better with European bees. 

Egg laying rates for queens evaluated in the daily egg laying 
experiment were compared with estimated egg laying rates derived from 
the brood production experiment (Table 6-12). There was no correlation 
between daily egg laying rates with either the estimated daily egg 
laying rates for the first 12 days or the estimated daily egg laying 
rates for 17 days. Estimated daily egg laying rates at day 12 was 
significantly correlated to the estimated overall egg laying rates at 
day 17 (P<0.05). 

Adequate pollen and nectar resources were available during the 
experiment. Each colony stored pollen and had an overall weight gain. 
There was no significant difference between colonies with Africanized or 
European worker bees with respect to pollen stored or weight gained. 
The amount of pollen stored by each colony was not significantly 
correlated with colony weight gain, total brood produced, or mean 
estimated daily mortality. Colony weight gain was positively correlated 
(P<0.05) with total brood produced and negatively correlated (P<0.05) 
with mean estimated daily mortality. 

Discussion 
The purpose of these experiments was to compare egg laying rates 
between Africanized and European queen honey bees and to determine if 
Africanized and European worker bees differentially affect brood 
production and/or the queen's egg laying behavior. Results from daily 
egg laying rates indicate that there was no significant difference 
between Africanized and European queens during the initial colony growth 
period under identical experimental conditions in Venezuela. There was 



103 
also no significant difference between Africanized and European worker 
bees on brood production rates or egg laying rates of either Africanized 
or European queens. 

Differences in egg laying and brood production rates could be 
evaluated by eliminating potential differences in foraging success 
between Africanized and European bees by providing surplus honey to each 
experimental colony and by controlling comb cell size. Differences in 
brood production caused by differences in resource utilization 
efficiency as a result of either smaller bee size or increased foraging 
success were, therefore, not evaluated. For a given amount of food, a 
greater number of smaller. Africanized bees can be produced compared 
with larger, European bees (Fletcher and Tribe 1977a; Tribe and Fletcher 
1977). However, this advantage for Africanized bees with respect to 
their smaller size, would only be present if resources were limited. 
Therefore, surplus honey was provided for each colony to reduce the 
effects of limited resources and the effects of differential foraging 
success between the two honey bee populations. Controlling comb cell 
size eliminated brood production differences based on bee size. For 
example, a given number of nurse bees may be able to rear more smaller 
bees than larger ones. In order that queen-worker bee interactions 
could be evaluated, European comb size was selected because of the 
difficulties both European queens and worker bees have with Africanized 
comb as discussed earlier. 

Egg laying rates observed during these experiments were within the 
range reported for Africanized bees in French Guiana (Winston and Taylor 
1980) but lower than the maximum reported for either population 
(Fletcher 1978; Ribbands 1953). Several factors can account for this 



104 
difference. First, the experimental colonies used in this study may 
have been only one-tenth the size of the managed, production colonies in 
which the maximum rates were observed. Because egg laying rates are 
correlated with the number of worker bees in the colony (Moeller 1958), 
the lower egg laying rates may have been a result of smaller colonies. 
Second, the intensity of colony disturbance, particularly during the 
daily egg laying experiment, would have reduced egg laying rates and 
increased egg and larval mortality. Third, egg laying rates for 
artificially inseminated queens may be lower than for naturally mated 
queens (Harbo and Szabo 1984). The purpose of this study was not to 
determine absolute egg laying rates but to compare egg laying and brood 
production rates for Africanized and European queens during the initial 
colony growth phase under identical conditions. 

There were no differences in egg laying and brood production rates 
for Africanized and European bees during the initial colony growth 
phase. When colonies increase in size and approach their maximum growth 
phase and queens are maximally challenged, there may be a difference 
between Africanized and European queens and colonies. However, 
comparisons of queen pupal weights (as a correlate of egg laying rates) 
suggest that there would be no difference between Africanized and 
European queens with respect to potential egg laying capacity (Chapter 
V). 

Colony growth rates and therefore reproductive rates are affected 
by two other demographic parameters: adult longevity and brood 
mortality. Winston and Katz (1981) found that European worker bees were 
longer lived than Africanized worker bees under identical conditions in 
Venezuela (26.3 compared with 22.7 days). This difference would give 



105 
European colonies a growth rate advantage. Unfortunately, brood 
mortality for both Africanized and European honey bees during the 
initial colony growth phase has not been investigated under identical 
conditions. 

Based on the results from these studies and those evaluating other 
colony demographic parameters, e.g., worker development times (Chapter 
II) and queen maturation rates (Chapter IV), it must be concluded that 
differences in reproductive rates between Africanized and European honey 
bees in South America cannot be attributed to intrinsic demographic 
factors. Reproductive rates in honey bees are a function of at least 
two other factors — resource availability and resource utilization 
efficiency. Chapter VIII presents a hypothesis to explain the success 
of Africanized honey bees based on differences In resource utilization 
efficiency. This hypothesis 1s based on differences between Africanized 
and European bees with respect to foraging behavior, brood production 
efficiency as a function of bee size, and resource- induced absconding. 



106 



TABLE 6-1. Daily egg laying rates of Africanized and European queens 
with European nurse bees, trial 1. 



ONE-DAY HIGH 



MEAN + SD (n) 



AFRICANIZED QUEENS 
A57 (W42) 
A62 (W81) 

EUROPEAN QUEENS 
YK (Y4) 
YDS (Y42) 
GK30 (Y63) 
GK30 (Y64) 

ANALYSIS^ 



550 
857 

951 
763 
710 
705 



473.2 + 68.5 (6) 

825.0 + 47.0 (3) 

922.0 + 41.0 (2) 

689.6 + 66.4 (5) 

636.7 + 80.8 (3) 
575.6 1 101.3 (5) 



NS 



^ann-Whitney U test, two-tailed, alpha = 0.05; evaluated for all 
samples (n = 24) . 



107 



TABLE 6-2. Daily egg laying rates of Africanized and European queens 
with European nurse bees, trial 2. 



ONE-DAY HIGH 



MEAN + SD (n) 



AFRICANIZED QUEENS 
AS7 (W42) 
A62 (W81) 

EUROPEAN QUEENS 
GK30 (Y63) 



ANALYSIS^ 



740 
953 

779 



699.8 +39.5 (4) 
915.8 +30.7 (4) 

736.8 +32.2 (4) 



NS 



^4ann-Whitney U test, two-tailed, alpha = 0.05; evaluated for all 
samples (n = 12) . 



TABLE 6-3. Daily egg laying rates of Africanized and European 
queens with Africanized nurse bees, trial 2. 



108 





QUEENS 


ONE 


-DAY 


HIGH 


MEAN + SD (n) 


AFRICANIZED 










A61 (W61) 






847 




799. S + 67.2 (2) 


AST (W41) 






725 




694.0 + 39.1 (4) 


A62 (W85) 






883 




798.7 + 92.3 (3) 


EUROPEAN QUEENS 










GK30 (Y61) 






850 




823.5 + 37.0 (4) 


YDS (YS2) 






949 




782.7 +231.7 (3) 


ANALYSIS^ 










NS 



^ann-Whitney U test, two-tailed, alpha = O.OS; evaluated 
for all samples (n = 16). 



TABLE 6-4. Effect of nurse bee genotypes on the daily 

egg laying rates of Africanized and European 
queens, trial 2. 



109 



MEAN + SD (n) 



AFRICANIZED NURSE BEES 

AFRICANIZED QUEENS 

A61 (W61) 

AST (W41) 

A62 (W85) 

COMBINED 

EUROPEAN QUEENS 

GK30 (Y61) 

YDS (Y52) 

COMBINED 

COMBINED AFRICANIZED 
AND EUROPEAN QUEENS 

EUROPEAN NURSE BEES 

AFRICANIZED QUEENS 

AS7 (W42) 

A62 (W81) 

COMBINED 
EUROPEAN QUEEN 

GK30 (Y63) 

COMBINED AFRICANIZED 
AND EUROPEAN QUEENS 



799.5 + 67.2 (2) 

694.0 + 39.1 (4) 

798.7 + 92.3 (3) 

752.3 + 79.6 (9) (A) 

823.5 + 37.0 (4) 

782.7 + 231.7 (3) 

806.0 + 138.0 (7) (B) 



775.8 + 108.4 (16) 



736.8 + 32.2 (4) 



(C) 



699.8 + 39.5 (4) 
915.8 + 30.7 (4) 
807.8 + 120.0 (8) (D) 



(E) 



784.1 + 103.3 (12) (F) 



110 



TABLE 6-5. Analyses of the effect of queen genotypes and nurse bee 

genotypes on the daily egg laying rates of Africanized and 
European honey bees. Letters represent the different 
treatments presented in Table 6-4. Mann-Whitney U test, 
two-tailed, alpha = 0.05 (Siegel 1956). 

HI: There is no difference in egg laying rates between Africanized 

and European queens. 

i. Africanized nurse bees 

A X B NS 

ii. European nurse bees 

D X E NS 

iii. Africanized and European 
nurse bees combined 

(A + D) X (B + E) NS 



H2: There is no differential effect of Africanized and European 
worker bees on egg laying rates. 

i. Africanized queens 

A X D NS 

ii. European queens 

B X E NS 

iii. Africanized and European 
queens combined 

C X F NS 



Ill 



TABLE 6-6. Daily egg laying rates of Africanized and European queens; 
comparison between genotypes within each population. 



MEAN + SD (n) ANALYSES 

AFRICANIZED QUEENS 

1. TRIAL 1— EUROPEAN NURSE BEES 

A57 (W42) 473.2 + 68.5 (6) 

A62 (W81) 825.0 + 47.0 (3) 

2. TRIAL 2— EUROPEAN NURSE BEES 

A57 (W42) 699.8 + 39.5 (4) 

A62 (W81) 915.8 + 30.7 (4) 

3. TRIAL 2— AFRICANIZED NURSE BEES 

A61 (W61) 799.5 + 67.2 (2) 

A57 (W41) 694.0 + 39.1 (4) 

A62 (W85) 798.7 + 92.3 (3) 



P<0.05 ^ 



P<0.05 ^ 



NS"^ 



4. 



COMBINING #2 AND #3 p<o.o5 b 

EUROPEAN QUEENS 

1. TRIAL 1— EUROPEAN NURSE BEES 

YK (Y4) 922.0 + 41.0 (2) *= 

YDS (Y42) 689.6 + 66.4 (5) 

GK30 (Y63) 636.7 + 80.8 (3) 

GK30 (Y64) 575.6 +101.3 (5) 

NS^ 

2. TRIAL 2~EUR0PEAN NURSE BEES 

GK30 (Y63) 736.8 + 32.2 (4) 

3. TRIAL 2— AFRICANIZED NURSE BEES 

GK30 (Y61) 823.5 + 37.0 (4) 

YD5 (Y52) 782.7 +231.7 (3) 

NS^ 

4. COMBINING #2 AND #3 IMS^ 



^ann-Whitney U test, two-tailed, alpha = 0.05. 

^Kruskal-Wallis one-way analysis of variance by ranks, alpha = 0.05. 
YK (Y4) was significantly different, P<0.05, from other genotypes in 
the group when evaluated by pairs using the Mann-Whitney U test, 
two-tailed, alpha = 0.05. 



112 

TABLE 6-7. Comparison of daily egg laying rates for sister queens with 
Africanized and European nurse bees. Mean + SD (n = sample 
size) . 



AFRICANIZED EUROPEAN 

NURSE BEES NURSE BEES ANALYSES^ 



AFRICANIZED QUEENS 

A57 (W41) 694.0 +39.1 (4) NS 

A57 (W42) 699.8+39.5 (4) 



A62 (W85) 798.7 +92.3 (3) NS 

A62 (W81) 915.8 +30.7 (4) 

EUROPEAN QUEENS 

GK30 (Y61) 823.5 +37.0 (4) NS 

GK30 (Y63) 736.8 +32.2 (4) 



^lann-Whitney U test, two-tailed, alpha = 0.05. 



113 



TABLE 6-8. Comparison of brood production in cm^ for Africanized and 

European queens during two periods of the first brood cycle. 

DAY 12 DAY 17 



AFRICANIZED NURSE BEES 
AFRICANIZED QUEENS 



USB^ Se'^ TB'= USB SB TB 



A62 (W81) 


1053 


962 


2015 


1484 


1301 


2785 


A57 (W42) 


416 


666 


1082 


493 


818 


1311 


EUROPEAN QUEEN 














GK30 (Y63) 


1775 


819 


2594 


1330 


1756 


3086 


EUROPEAN NURSE BEES 














AFRICANIZED QUEENS 














A62 (W85) 


906 


823 


1729 


434 


1354 


1788 


A57 (W41) 


1481 


487 


1968 


1018 


1256 


2274 


EUROPEAN QUEEN 














GK30 (Y61) 


1614 


728 


2342 


1580 


1519 


3099 



ANALYSES^ 

AFR WORKERS x EUR WORKERS NS NS NS NS NS NS 



^USB = unsealed brood (eggs and larvae). 

SB = sealed brood (pre-pupae and pupae). 
^TB = total brood. 
°Mann-Whitney U test, one-tailed, alpha = 0.05. AFR = Africanized; 

EUR = European. 



114 



TABLE 6-9. Colony adult population changes during first brood cycle. 



AFRICANIZED NURSE BEES 
AFRICANIZED QUEENS 



ESTIMATED DAILY 
DAY 0^ DAY 12'=' DAY 17^ MORTALITY RATE^ 



A62 (W81) 
A57 (W42) 


10,169 
10,169 


4,850 
4,408 


2,634 
2,007 




443 
480 


EUROPEAN QUEEN 












GK30 (Y63) 


10,169 


6,511 


4,987 


MEAN 


305 


EUROPEAN NURSE BEES 


= 409 


AFRICANIZED QUEENS 












A62 (W85) 
AST (W41) 


10,435 
10,435 


5,375 
5,102 


3,267 
4,297 




422 
361 


EUROPEAN QUEEN 












GK30 (Y61) 


10,435 


6,321 


4,607 


MEAN 


343 




= 375 


ANALYSIS® 










NS 



^Population at Day was estimated as described in methods. 

■^Population at Day 12 estimated by: Day - [(Day - Day 17) (12/17)], 

^Population at Day 17 was estimated as described in methods. 

^(Day - Day 17)/17. 

®Mann-Whitney U test, one-tailed, alpha = 0.05. 



115 



TABLE 6-10. Brood production during first brood cycle expressed as a 
percent of adult population. 



DAY 12 DAY 17 



Db 



USB^ SB^ TB«^ USB SB TB 



AFRICANIZED NURSE BEES ~ 

AFRICANIZED QUEENS 

A62 (WSl) 92.3 84.3 176.6 239.4 209.9 449.3 

A57 (W42) 40.1 64.2 104.3 104.4 173.2 274.6 

EUROPEAN QUEEN 

GK30 (Y63) 115.9 53.4 169.3 113.3 149.6 262.9 

EUROPEAN NURSE BEES 

AFRICANIZED QUEENS 



A62 (W85) 
A57 (W41) 


71.6 
103.2 


65.1 
33.9 


136.7 
137.1 


56.4 
100.7 


176.1 
124.2 


232.5 
224.9 


EUROPEAN QUEEN 














GK30 (Y61) 


108.5 


48.9 


157.4 


145.8 


140.1 


285.9 


ANALYSES^ 














AFR WORKERS x EUR WORKERS 


NS 


NS 


NS 


NS 


NS 


NS 



Unsealed brood (eggs and larvae); colony adult population estimated 
for Day 12 and Day 17 as described in methods. 
Sealed brood (pre-pupae and pupae). 
^Total brood. 

Mann-Whitney U test, one-tailed, alpha = 0.05. AFR = Africanized; 
EUR = European. 



116 



TABLE 5-11. Comparison of overall egg laying rates 
between Africanized and European sister 
queens with Africanized and European 
nurse bees during the first brood cycle. 



AFRICANIZED QUEENS 



EUROPEAN QUEENS 



AFRICANIZED EUROPEAN 
NURSE BEES^ NURSE BEES 



A62 (W81) 696 

A62 (W85) — 447 

A57 (W42) 328 

A57 (W41) — 558 



GK30 (Y63) 772 

GK30 (Y61) — 775 



^Overall egg laying rate = (.TB-^-j/M) i4,25) ^ 



117 



Table 6-12„ 


Comparison of egg 


laying 


rates 


during daily eqq 


layi 




rate experiment and brood 


production 


experiment. 








AFRICANIZED 






EUROPEAN 




QUEENS 




NURSE BEES 






NURSE BEES 




ELR^ 


BP12^ 


BP17= 


ELR 


BP12 


BP17 


AFRICANIZED 
















A62 (W81) 


— 


714 


696 




916 






A62 (W85) 


799 


— 


— 






612 


447 


A57 (W42) 


— _ 


383 


328 




700 






A57 (W41) 


694 


— 


— 






697 


568 


EUROPEAN 
















GK (Y63) 


__ 


918 


772 




737 






GK (Y61) 


824 


~ 


— 






829 


775 


ANALYSES^ 


ELR 

ELR 

BP12 


X BP12 
X BP17 
X BP17 


NS 
NS 
P<0.05 











^Daily egg laying rate experiment, means. 

Brood production experiment; egg laying rate estimated for first 
12 days of brood cycle by dividing total brood at day 12 by 12. 
Brood production experiment; egg laying rate estimated for first 
17 days of brood cycle by dividing total brood at day 17 by 17. 
Spearman rank correlation, alpha = 0.05; Africanized and 
European nurse bees combined. 



CHAPTER VII 
SUCCESSFUL HYBRIDIZATION BETWEEN AFRICANIZED AND EUROPEAN HONEY BEES 
IN VENEZUELA WITH IMPLICATIONS FOR NORTH AMERICA 

Introduction 
In 1956 African honey bees. Apis mellifera scutellata . formerly 
classified as A. m. adansonii (Ruttner 1976a, 1976b, 1981), were 
imported Into southeastern Brazil (Kerr 1967). Their hybridized 
descendents, known as Africanized honey bees (Goncalves 1982), have 
rapidly spread throughout tropical South and Central America as far 
north as Honduras and El Salvador (Rinderer 1986). The dispersion from 
their original importation site into new areas has been rapid— 200-500 
km per year (Taylor 1977, 1985; Winston 1979a). As Africanized honey 
bees have dispersed into new areas, they have rapidly increased in 
number (Otis 1982a) and have attained dramatic population densities 
(Michener 1975): 4-8 colonies/km^ (Taylor 1985), or as high as 107.5 
colonies/km^ in the cerrado habitats in the Brazilian states of Goias 
and Mato Grosso (Kerr 1971 cited in Michener 1975). There are now 
probably more than ten million feral colonies of Africanized honey bees 
in South and Central America (Winston, Taylor and Otis 1983). Their 
success in these habitats, compared with European populations of honey 
bees, may be attributed to their foraging behavior which is more suited 
to the resource patterns of the tropics (Nunez 1973, 1979a, 1982; 
Rinderer, Bolten, Collins and Harbo 1984; Rinderer, Collins and Tucker 
1985; Winston and Katz 1982). As a result of both their foraging 

118 



119 
success and length of time throughout the year that resources are 
available in the tropics. Africanized honey bees have a high annual 
reproductive rate, which is responsible for both their rate of dispersal 
into new areas and high colony densities. Net reproductive rates for 
Africanized bees have been estimated to be 16 colonies per colony per 
year based on demographic data collected in French Guiana (Otis 1980, 
1982a) compared with 0.92-0.96 (Seeley 1978) or 3-3.6 when afterswarms 
are counted (Winston 1980a; Winston, Taylor and Otis 1983) for European 
honey bees in North America. 

Particularly in the region of their importation (southeastern 
Brazil), there has been ample opportunity for hybridization with both 
managed and feral European honey bees [primarily A. m. mellifera and A. 
m. li gust tea which had been imported into Brazil by 1845 (Gerstaker 
cited in Pellet 1938; Woyke 1969)]. However, despite opportunity for 
hybridization. Africanized honey bees have maintained behavioral, 
chemical and morphological characteristics similar to their African 
parental population and distinguishable from European honey bees: 
colony defense behavior (=stinging behavior) (Collins, Rinderer, Harbo 
and Bolten 1982; Stort 1974, 1975a, 1975b, 1975c, 1976); reproductive 
rates (Fletcher 1978; Fletcher and Tribe 1977a; Otis 1980, 1982a); 
absconding behavior (reviewed by Fletcher 1978; Winston, Otis and Taylor 
1979; Winston, Taylor,and Otis 1983); foraging and hoarding behavior 
(Nunez 1973, 1979a, 1982; Rinderer, Bolten, Collins and Harbo 1984; 
Rinderer, Bolten, Harbo and Collins 1982; Rinderer, Collins and Tucker 
1985; Winston and Katz 1982); worker bee longevity (Winston and Katz 
1981); development times (Chapters II and IV; Harbo, Bolten, Rinderer 
and Collins 1981); selection preferences for nest cavity sizes 



120 
(Rinderer, Collins, Bolten and Harbo 1981; Rinderer, Tucker and Collins 
1982); allozyme patterns (Nunamaker and Wilson 1981; Sylvester 1982); 
cuticular hydrocarbon composition (Carlson and Bolten 1984, and 
unpublished data); adult bee size and comb cell size (Chapter III; 
Michener 1975); and morphometric relationships (Daly and Balling 1978). 

This apparent lack of evidence for hybridization has been 
attributed primarily to some degree of reproductive isolation between 
the Africanized and European populations (Kerr and Bueno 1970; Taylor 
1985). Three isolating mechanisms have been suggested: assortative 
mating (Kerr and Bueno 1970); physiological incompatibility with respect 
to the drone ejaculation response (Kerr and Bueno 1970); and differences 
in drone and presumably the queen flight times between the Africanized 
and European populations (Taylor 1985; Taylor, Kingsolver and Otis in 
press). At best, these mechanisms may be only partially effective and 
are not likely to account for the apparent lack of hybridization. For 
example, Kerr and Bueno (1970) present data to support assortative 
mating even though 35% and 42% of the matings evaluated were hybrid. 
With respect to differences in queen and drone flight times, data from 
Venezuela suggest that mean peak drone flight times for Africanized and 
European populations are separated by only 23 minutes and that drones 
from both populations are present in the mating areas at all times 
during the approximately three hour flight period (Taylor, Kingsolver 
and Otis in press). 

A more probable argument for the maintenance of the African 
characteristics is based on selection advantages for the African 
genotype in tropical habitats of South America, which are characterized 
by resource distribution patterns similar to the ones in Africa where 



121 
these bees evolved. Two lines of evidence support this selectionist 
argument. First, the foraging behavior of the Africanized honey bees, 
characterized by solitary foraging and less colony recruitment, would be 
more adaptive in tropical areas with rich, but dispersed, resources 
(Rinderer, Bolten, Collins and Harbo 1984; Rinderer, Collins and Tucker 
1985). Second, there is much historic evidence that European honey bees 
have not been successfully maintained (probably due to starvation) in 
many areas of tropical South America that are now densely populated with 
Africanized honey bees (Bolten, personal observation; Michener 1972; 
Winston, Taylor and Otis 1983). 

The selectionist argument allows for hybridization between 
Africanized and European populations, with the African genotype being 
selected for under the physical and biotic parameters characteristic of 
tropical areas. Selection for the African genotype would then account 
for the present population in South and Central America being 
behaviorally, chemically and morphologically similar to the African 
parental type. 

These two alternative hypotheses for the maintenance of the African 
parental characteristics— reproductive isolation between the Africanized 
and European honey bee populations versus selection for the African 
genotype in tropical regions— suggest different scenarios for the 
potential impact of Africanized honey bees in North America. One 
scenario resulting from reproductive isolation would limit Africanized 
honey bees in their northern movement because of their inability to 
overwinter (Taylor 1985; Taylor and Spivak 1984). This is based on the 
observation that the parental African population as well as Africanized 
bees do not have the thermoregulatory capabilities to survive the cold 



122 
temperatures of temperate winters (Nunez 1979b; Woyke 1973). The other 
scenario based on successful hybridization and resultant genetic 
introgression would have the stinging behavior of the Africanized honey 
bees and the potential public health hazard widespread in North America 
and not limited to the southern regions. 

A comparison between hybrid and non-hybrid mating success is 
lacking for the Africanized and European populations of honey bees. 
Although reciprocal Fl crosses (Africanized queen x European drone; 
European queen x Africanized drone) can be successfully produced with 
artificial insemination (Chapater III), natural matings have been 
difficult to evaluate because of the inability to distinguish between 
hybrid and non-hybrid progeny on the individual level. It is not 
possible to detennine which drones have mated a queen because honey bees 
mate in the air away from the hive, and there are no available genetic 
markers in the Africanized population. Identifying Individual honey 
bees as either Africanized or European is not presently possible 
(Carlson and Bolten 1984; Page and Erickson 1985; Rinderer and Sylvester 
1981). However, there is promise that DNA analyses will allow 
individuals to be identified (Hall in press). 

This paper evaluates an important parameter with respect to the 
question of successful hybridization between Africanized and European 
honey bees: the mating success of both Africanized and European queens 
with Africanized drones. The mating success of European queens x 
Africanized drones was compared with the mating success of Africanized 
queens x Africanized drones in an isolated area in Venezuela with no 
European honey bees but with a feral population of Africanized honey 
bees. Because of present legal restrictions. Africanized bees cannot be 



123 
taken Into areas with only European honey bees so the reciprocal cross 
was not evaluated. However, evaluating the success of the European 
queen x Africanized drone cross is important because it represents the 
most probable initial hybridization that will occur when Africanized 
honey bees invade North America (Mexico and U.S.A.). 

Methods 

The Africanized queen mother (A26) was removed from a feral colony 
in eastern Venezuela where there were no known European honey bees. The 
colony was identified as Africanized by its behavior and small comb cell 
size characteristic of the Africanized population (4.5-5.0 mm between 
opposite sides of the hexagon. Chapter III). The European queen mother 
(L13) was produced by a commercial queen producer in the southeastern 
U.S.A. and shipped to Venezuela. 

Experimental queens were produced from each queen mother by the 
standard queen rearing technique of transferring (grafting) 12 to 18- 
hour-old larvae into artificial queen cells that were then introduced 
into queen-cell producing colonies (Laidlaw 1979). After the queen 
cells were sealed, each cell was protected by a wire mesh cylinder (mesh 
size = 3.0 mm). Virgin queens were allowed to emerge in the cell- 
producing colonies. Newly emerged virgins were marked for 
identification and then stored in a strong, queenless colony. The 
following day they were introduced into individual, five-frame colonies 
with Africanized worker bees from which natural matings could occur. 
Queens were released into these mating colonies using standard three- 
hole mailing cages (Laidlaw 1979). Based on earlier calculations, 
natural release from the mailing cages was estimated to take 2.5-3.0 
days. Marked, virgin queens were introduced into mating colonies rather 



124 
than mature queen cells so that the Identity of the experimental queens 
would later be certain. The mating colonies were located 1n an area In 
eastern Venezuela where there was no known European honey bees, but 
which was densely populated with feral Africanized colonies. 

Eighteen days after Introduction into mating colonies, the queens 
were collected, and the spermatazoa in their spermatheca were counted 
using hemacytometers (Mackensen and Roberts 1948; Mackensen and Tucker 
1970). The criterion for mating success was the number of spermatozoa 
in the spermatheca. In addition, the age of the queen when oviposition 
first started was calculated by determining the age of the oldest brood 
in each colony. 

Results 
Numbers of spermatozoa counted in the spermatheca of Africanized 
and European queens are summarized in Table 7-1. The mean number of 
spermatozoa in Africanized queens (4.09 +0.50 million) was not 
different from the mean number in European queens (4.12 +0.58 million; 
t-test, two-tailed, alpha = 0.05). European queens began oviposition on 
the 10th day post-emergence, one day sooner than the Africanized queens 
(P<.001; Kolmogorov-Smirnov two-tailed test, chl-square distribution, df 
= 2). The time from adult emergence to initiation of oviposition 
reported here (Table 7-1) Is longer than the maturation Interval 
reported in Chapter IV, which may be a result of having introduced 
virgin queens into the mating colonies rather than mature queen cells. 
There was no correlation between the time post-emergence to initiation 
of oviposition and the number of spermatozoa in the spermatheca 
(Spearman's rank correlation coefficient, two-tailed, alpha = 0.05). 
The acceptance of the Africanized and the European virgin queens 



125 
introduced Into the mating nuclei was 50% and 61%, respectively. There 
was no significant difference in acceptance (Fisher's exact probability 
test, alpha = 0.05) . 

Discussion 
Evidence for Hybridization 

There appears to be no effective reproductive isolating mechanism 
operating between the Africanized and European populations. The mating 
of both Africanized and European queens with Africanized drones was 
equally successful as judged by the number of spermatozoa in the 
spermatheca. Offspring from the hybrid crosses were viable with no 
apparent difference in mortality as determined by the uniformity of the 
brood pattern. Kerr and Bueno (1970) report that there may be a 
difference in ejaculation response between Africanized and European 
drones that may provide a potential isolating mechanism. Even if this 
exists, European queens were still able to successfully mate with 
Africanized drones without any apparent problem, as determined by both 
the numbers of spermatozoa in their spermatheca and the age when 
oviposit ion began. 

Although the same queen pheromone is produced by three sympatric 
Asiatic species of Apis (Butler, Calam and Callow 1967; Shearer, Boch, 
Morse and Laigo 1970), reproductive isolation occurs between the three 
species because there is no overlap in times of drone flight (Koeniger 
and Wijayagunasekera 1976). The situation between Africanized and 
European populations of honey bees is quite different with respect to 
the time of flight of the queens and drones. Data from Venezuela show 
that flight times for Africanized and European drones completely overlap 
during the approximately three hours of mating flight activity with only 



126 
23 minutes separating the mean times of peak flight activity for each 
population (Taylor, Kingsolver and Otis in press). This difference does 
not provide a satisfactory mechanism for reproductive isolation between 
the Africanized and European honey bee populations—particularly because 
any unfavorable climatic conditions, e.g., high winds, cloudiness, high 
humidity, or rain (Gary 1975), cause mating flights of reproductives 
from both populations to more completely converge to times of favorable 
weather conditions. The data on reproductive success (determined by the 
number of spermatozoa in the spennatheca) of European queens mating with 
Africanized drones presented in this study demonstrate that any 
differences in mean peak flight times did not effectively prevent 
hybridization of European queens with Africanized drones. 

Evidence that extensive hybridization has already occurred between 
the introduced African honey bees and the previously established 
European honey bees can be demonstrated by the increase in genetic 
diversity in the Africanized population. For example, Adams, Rothman, 
Kerr and Paulino (1977) concluded that the large increase in number of 
sex alleles in the population of honey bees in southeastern Brazil is a 
result of hybridization between African and European honey bees. Page 
and Erickson (1985) also suggest that hybridization has occurred based 
on the variation in behavior and appearance of Africanized bees in 
Venezuela. 
Impact of Hvbridizatinn 

Demonstration that successful hybridization can and does occur has 
important implications with respect to the potential impact the 
Africanized honey bee will have in North America. First is the negative 
impact that hybridization would have. The parental African population 



127 
as well as Africanized bees may not have the thermoregulatory 
capabilities to survive the cold temperatures of temperate winters 
(Nunez 1979b; Woyke 1973). Based on the temperature limits of the 
parental population, Taylor (1985) and Taylor and Spivak (1984) 
predicted the northern limits of Africanized honey bees in North 
America. However, Africanized honey bees may acquire, through 
hybridization with European honey bees in Mexico and southern U.S.A., 
the ability to overwinter farther north than is presently expected. 
That is, the overwintering genome of the European honey bees may become 
introgressed into the Africanized genome. Or, the corollary, the 
stinging behavior characteristic of the Africanized honey bees may 
become introgressed into the overwintering European population. 
Successful genetic introgression of these traits may not be a rapid 
process because these traits are polygenic and/or may involve coadapted 
genomes. However, because hybridization occurs, the potential for 
successful genetic introgression exists and must be considered. 
Hybridization may, therefore, result in the stinging behavior of the 
Africanized honey bees becoming a potential public health hazard 
throughout North America, not just in the warmer southern regions. That 
this may be the unfortunate outcome of hybridization is supported by 
recent investigations in Argentina, which have demonstrated that 
Africanized honey bees are distributed farther south than predicted 
based on temperature limits of the parental population (Dietz, Krell and 
Eischen 1985; Krell, Dietz and Eischen 1985). 

The U.S. Department of Agriculture Economic Research Service has 
recently evaluated the potential impact of Africanized honey bees in the 
U.S.A. (McDowell 1984). Unfortunately this report does not consider the 



128 
possibility that further hybridization between Africanized and European 
honey bees might result in the stinging behavior of Africanized honey 
bees becoming established throughout the northern regions of the U.S.A. 
The economic consequences as j^ell as public health hazards may be more 
widespread throughout North America than previously thought. For any 
solutions to the Africanized honey bee problem to be successful, a 
realistic assessment of the potential problem is necessary. 

However, hybridization may also have a positive impact. Coupled 
with selection favoring both the foraging and thermoregulatory behavior 
of European honey bees in temperate regions, hybridization between 
Africanized and European bees may have the positive effect of increasing 
the rate at which African genes become rare in the population. There is 
a large population of European honey bees, both managed and feral, in 
North America (perhaps greater than IS million colonies in the U.S.A. 
alone), which is particularly dense in the south where the Africanized 
bees will first enter the U.S.A. The invading Africanized population 
would be quite small relative to the existing European population, 
increasing the frequency of hybridization and resultant "swamping" (or 
diluting) of African genes. 

The problem of Africanized honey bees may be reduced prior to their 
entry into the U.S.A. because of both the potential for hybridization 
and competition for available floral resources with European honey bees 
in Mexico. Mexico has a larger population of European honey bees than 
any other country in Latin America. When Africanized honey bees enter 
Mexico, they will be entering a region that already has an extensive, 
established population of European honey bees, both managed and feral 
[2.6 million managed colonies alone (Zozaya cited in Taylor 1985)]. 



129 
Competition with an established population of honey bees for limited 
floral nectar and pollen resources will be much greater than Africanized 
bees have previously experienced In any areas 1n South America. This 
competition will greatly slow their dispersal. In many areas of Mexico 
and southern U.S.A., pollen and nectar resources are already close to 
being saturated by the existing honey bee population. In addition, 
under temperate resource conditions, the foraging behavior of 
Africanized honey bees (which is more appropriate to tropical resource 
patterns) will be at a disadvantage relative to the foraging behavior of 
the European honey bee population, which is characterized by greater 
colony recruitment (Rinderer, Bolten, Collins and Harbo 1984; Rinderer, 
Collins and Tucker 1985; Visscher and Seeley 1982). 

The selective advantage of the foraging and/or thermoregulatory 
behavior of European honey bees has been demonstrated in temperate 
regions. African honey bee queens were introduced into North America 
during the late 1800's and early 1900's when the beekeeping industry In 
the U.S.A. was developing (Morse et al. 1973) and more recently, into 
Louisiana (Cantwell 1974; Morse et al. 1973; Taber 1961). However, due 
to hybridization and selection against the African genotype, the impact 
of these introductions of African bees is undetectable today. There 
have also been unsuccessful introductions of African and Africanized 
bees into Europe (Cantwell 1974; Morse et al. 1973; Woyke 1973). 
Therefore, these factors— a large, established population of European 
honey bees, and both a foraging and thermoregulatory behavior in 
European bees better adapted to temperate conditions— precludes using 
South and Central America as a model for North America in predicting the 
impact, as well as the rate of spread, of Africanized honey bees. 



130 
K1n recognition has been suggested as another mechanism that may 
help preserve the African genotype in the hybridized honey bee 
population in South and Central America (Hall in press). However, data 
presented in Table 4-2 demonstrate that this is unlikely because both 
Africanized and European worker bees reared both Africanized and 
European queens with equal frequency. 

There have been several suggestions that Africanized drones may 
have a mating advantage over European drones (Kerr and Bueno 1970; 
Michener 1975; Morse 1984; Rinderer 1986; Rinderer, Hellmich, Danka and 
Collins 1985; Taylor 1985). Taylor (1985) suggests that this mating 
advantage would reduce hybridization. On the contrary, a mating 
advantage for Africanized drones would increase the rate of 
hybridization between Africanized and European honey bee populations. 
When Africanized honey bees begin invading North America (Mexico and 
U.S.A.), they will be greatly outnumbered by the established European 
honey bee population. With a mating advantage, the frequency of 
European queen x Africanized drone matings would be greater than 
expected based solely on the relative frequency of each population, 
thereby resulting in greater hybridization. 

Unfortunately, Africanized honey bee research has been 
characterized by conceptualizing the Africanized honey bee as a distinct 
entity that is reproductively isolated (Taylor 1977, 1985; Taylor and 
Spivak 1984) rather than as a population within a species fully capable 
of hybridization. Clearly, the Africanized honey bee is not a species 
invading a new habitat (North America) that is free of competition from 
conspecifics. Thus, the spread of Africanized bees (African genes) in 
temperate North America will be: 1) farther north than predicted by the 



131 
geographic limits of the parental population because of hybridization 
and resultant genetic int regression; 2) slowed considerably by 
competition for available resources by an established population of 
European honey bees; 3) swamped through hybridization with a more 
numerous, established population of European honey bees; 4) at a 
disadvantage with respect to foraging behavior; and 5) limited by 
selection against those colonies that have not acquired through 
hybridization the ability to overwinter. 

There has been a general lack of support for the selectionist 
argument for the maintenance of the African genotype in favor of the 
hypothesis of reproductive isolation. With the demonstration that 
hybridization is successful, coupled with the recent observations of the 
distribution patterns of Africanized honey bees in Argentina (Dietz, 
Krell and Eischen 1985; Krell, Dietz and Eischen 1985), further 
consideration of the selectionist argument for the maintenance of the 
African characteristics in the Africanized honey bee in South and 
Central America is necessary. 



132 
TABLE 7-1. Mating success of Africanized and European honey bee que 



jeens. 



TIME TO NO. SPERM 

OVIPOSITION^ (x 10^)^ 



CORRELATIONS^ 



AFRICANIZED u 4.09+0.50 

GENOTYPE (A26) (8) (8) 

EUROPEAN 10 4.12+0.58 

GENOTYPE (L13) (H) (n) 



NS 
NS 



ANALYSES P<0.00ld 



NS^ 



Median days post-emergence to initiation of oviposit ion (sample size) 
One-day-old virgins were introduced into mating colonies. 

^ean + SD (sample size) of spermathecal spermatozoa number. 

^Spearman's rank correlation coefficient, two-tailed, alpha = 0.05. 
Kolmogorov-Smirnov two-tailed test, chi-square distribution, df =2. 
t-test, two-tailed, alpha = 0.05. 



CHAPTER VIII 

DISCUSSION: FACTORS COI^RIBUTING TO THE SELECTION ADVANTAGE OF 

AFRICANIZED HONEY BEES IN SOUTH AMERICA— 

THE RESOURCE UTILIZATION EFFICIENCY HYPOTHESIS 

Success of TntrodiiCfid P opulatinns of Hnnfiv Bff.e^ 
Thirty years ago African honey bees, Ap1s melUfera scutellata 
[formerly classified as A. m. adansonll (Ruttner 1976a, 1976b, 1981)], 
were Introduced into southeastern Brazil (Kerr 1967). Offspring, known 
as Africanized honey bees because of hybridization with European honey 
bees (Goncalves 1982), have rapidly dispersed throughout South America, 
sometimes achieving dramatically high population densities (Michener 
1975; Taylor 1977, 1985). In 1982 Africanized honey bees entered Panama 
(Buchmann 1982) and by 1986 were as far north as Honduras and El 
Salvador (Rinderer 1986). The success and biological impact of 
Africanized honey bees in these tropical and sub-tropical regions, 
compared with the lack of success of European honey bees in these same 
regions, is a result of a selection advantage for the Africanized (= 
African) genotype in tropical resource and climatic conditions. The 
difference in success between Africanized and European honey bees is 
evidenced by the fact that 

European bees in Brazil were never commonly found living wild 
in the forests and countryside. This was especially true in 
tropical forest regions, where honey bees were virtually 
restricted to a few apiaries.. .Everyone questioned on the 
matter emphasized the increase in bees away from apiaries that 
occurred with the arrival of the Brazilian [Africanized] bees. 
(Michener 1972, p. 15). 



133 



134 
The selection advantage for Africanized bees may be a result of 
behavioral and/or physiological characteristics that may include 
differences in resource utilization and/or colony demography. It is not 
surprising that Africanized honey bees are better adapted to tropical 
conditions than are European honey bees, considering the former are 
derived from imported African bees that evolved under similar tropical 
and sub-tropical conditions in Africa. Fletcher (1978) has reviewed the 
biological characteristics of the parental population of African honey 
bees in Africa. 

The spread and impact of Africanized honey bees in South America 
must, however, be kept in perspective. European honey bees introduced 
into North America early in the 17th century (Pellett 1938) dispersed 
throughout North America, also achieving high population densities. In 
general, honey bees are very successful not only in their native 
habitats but in almost every region where they have been introduced. 
Their success is based on a highly developed social system that allows 
honey bees to: 1) develop large, perennial colonies that are able to 
buffer climatic changes; 2) efficiently utilize resources because of 
advanced communication and recruitment systems; and 3) defend against 
both vertebrate and invertebrate predators because of their very 
effective colony defense behavior. 

The question with which we are concerned in these studies is not 
what makes A. me] 1 if era more successful than other species nor what 
impact introduced honey bees have on native pollinator communities (see 
Roubik 1978, 1979, 1980, 1982, 1983; Roubik and Buchmann 1984). Nor is 
it a question of comparing Africanized honey bees in tropical regions 
with European honey bees in temperate regions (see Winston, Dropkin and 



135 
Taylor 1981 and Winston, Taylor and Otis 1983). Rather, the question 
is: what are the differences between Africanized and European 
populations of A. mell if era that make Africanized honey bees more 
successful in tropical regions not only in South America but also in 
Africa? European honey bees have not been successfully introduced into 
tropical areas of Africa despite numerous attempts (Fletcher 1977b, 
1978). 

Factors Affecting Honev Be e Reproductive Rates 
The success of Africanized honey bees in South America — as judged 
by their rate of dispersal and their population densities (Michener 
1975; Taylor 1977, 1985) — must surely be a result of a high reproductive 
rate. What are the differences between Africanized and European honey 
bees that allow for high reproductive rates in Africanized bees, and can 
these differences account for the impact of Africanized honey bees? 
This question does not involve identification or analysis of the 
proximal factors that are responsible for initiating reproduction, but 
does involve analysis of the components that affect the rate of 
reproduction. 

Reproductive rates in honey bees are a result of an interaction of 
at least three factors, all of which affect colony growth rates: 
resource availability, resource utilization efficiency (foraging 
success, brood production efficiency, and bee size), and colony 
demographic parameters (primarily queen fecundity and adult worker bee 
longevity). Therefore, in order to evaluate reproductive differences 
between Africanized and European honey bee populations, all three 
factors need to be considered. Unfortunately, early research evaluated 
reproductive rates of Africanized honey bees by comparing data for 



136 
Africanized honey bees from South America with data from studies of 
European honey bees from North America, which not only were collected 
under different resource conditions but also different experimental 
conditions (Otis 1980, 1982a; Winston 1979b, 1980a; Winston, Dropkin and 
Taylor 1981). 

Results from these earlier studies characterized the Africanized 
population as one with a dramatically high annual colony reproductive 
rate— four to five times greater than European honey bees in temperate 
regions (Otis 1980, 1982a; Winston 1980a; Winston, Taylor and Otis 
1983). However, because these comparisons were not based on data 
collected under similar environmental or experimental conditions, they 
are inappropriate comparisons and cannot be used to identify either the 
factors responsible for the difference in reproductive rates between the 
two populations or the factors responsible for the success of 
Africanized honey bees. Because the experimental conditions were 
different (e.g., hive volume), these comparisons were also inappropriate 
for comparing temperate and tropical honey bee populations. Were the 
apparent differences in colony reproduction between the two honey bee 
populations the result of differences in: 1) colony demography; 2) 
environmental and climatic factors; 3) experimental design; 4) resource 
utilization efficiencies; or 5) some combination of factors? Are there 
intrinsic differences between the two honey bee populations with respect 
to colony demographic parameters that allow for a more rapid colony 
growth rate and result in a greater reproductive rate for the 
Africanized honey bee population? Or, are the differences in 
reproductive rates a result of climatic conditions and/or resource 
availabilities and utilization in the tropics compared with temperate 



137 
regions? Finally, were the relatively high reproductive rates observed 
for Africanized honey bees in these studies (Otis 1980, 1982a) simply an 
artifact of experimental conditions, particularly with respect to brood- 
nest crowding? 

Brood-nest crowding is a primary stimulus for reproductive swarming 
in honey bees (Baird and Seeley 1983; Simpson 1966, 1973; Simpson and 
Riedel 1963). However, the experimental conditions affecting brood-nest 
crowding for Africanized colonies in South America were significantly 
different from the experimental conditions for European colonies in 
North America: the nest cavity volume for Africanized colonies was 22 
liters (Otis 1980; Winston 1979b) compared with 42 liters for European 
colonies in North America (Winston 1980a). Despite these problems with 
respect to making valid comparisons, the earlier studies (particularly 
Winston 1979b) leave one with the impression that the apparent 
differences in reproductive rates between the two populations were 
primarily due to differences in demographic parameters and not to 
differences in environmental and experimental conditions, resource 
utilization, or some combination of factors. 

What is the consequence of comparing reproductive rates of 
Africanized honey bees in South America with those of European honey 
bees in North America without considering differences in environmental 
conditions? Certainly, environmental conditions in temperate regions 
impose strict limits on the length of the reproductive (= swarming) 
season for honey bees because of a reduced growing season when floral 
resources (nectar and pollen) are available. The honey bee reproductive 
season is significantly shorter than the growing season, because 
colonies first have to grow to reproductive size before swarming can 



138 
occur. In addition, offspring (swarms) need a rather long period of 
time to grow and to hoard necessary surplus honey, while there are still 
floral resources available, in order to prepare for winter. 

Most mortality of honey bee colonies in temperate areas occurs 
primarily due to starvation during winter: 77% for first-year colonies 
and 90% for established colonies (Seeley 1978, 1983). There is a high 
energetic cost of maintaining proper brood nest or cluster temperature 
during the cold winter. In addition, the high energetic cost of winter 
survival in temperate regions may greatly reduce the survivorship of 
small, secondary swarms or afterswarms, which would greatly reduce the 
net reproductive rate of honey bees in temperate regions. In contrast, 
periods of resource dearth in the tropics are not only shorter, but 
require less stored honey (per unit time) to enable the colonies to 
survive because of reduced energetic costs for maintaining brood nest 
temperature. 

The reproductive season in French Guiana, South America, was nine 
months (Winston 1980b) compared with two to four months for North 
America (references cited in Winston 1980b and Winston, Dropkin and 
Taylor 1981). Is it a coincidence that the difference in the annual 
reproductive rate of Africanized bees in South America compared with 
European bees in North America, 16 vs. 3.0-3.6, respectively (Otis 
1982a; Winston 1980a; Winston, Taylor and Otis 1983), is approximately 
of the same order (factor of 4-5) as the difference in the length of the 
reproductive season between South America and many areas of North 
America? 

Another factor affecting differences in reproductive rates between 
honey bee populations in South America with those in North America is 



139 
the extent to which brood rearing ceases during resource dearths. 
European honey bee populations in temperate areas have a distinct 
seasonal decline in brood production and may stop brood rearing 
altogether for a variable period during winter (Bodenheimer 1937; 
Bodenheimer and Ben-Nerya 1937; Jeffree 1955; McLellan 1978; Nolan 1925, 
1928). On the other hand, Winston reports that many Africanized 
colonies in French Guiana 

persist during the relative dearth season (March to June) 
without the cessation of brood rearing characteristic for 
temperate conditions, and are strong enough (i.e., have a 
relatively high worker population and sufficient young 
workers) to grow rapidly to swarming strength when resources 
improve. (Winston 1980b, p. 164). 

This difference between tropical and temperate conditions allows 

tropical honey bee colonies to grow rapidly when resources become 

available and thereby increase their potential reproductive rates 

compared with temperate honey bee colonies. 

More recently, investigations of both Africanized and European 

honey bee populations under identical experimental conditions in 

Venezuela have been undertaken. These studies have evaluated both 

demographic parameters as well as resource utilization behaviors. These 

investigations include the research presented in Chapters II-VII; 

studies by Winston and Katz (1981, 1982); and the research by the U.S. 

Department of Agriculture Bee Breeding and Stock Center Laboratory 

(Collins, Rinderer, Harbo and Bolten 1982; Harbo, Bolten, Rinderer and 

Collins 1981; Rinderer, Bolten, Collins and Harbo 1984; Rinderer, 

Bolten, Harbo and Collins 1982; Rinderer, Collins, Bolten and Harbo 

1981; Rinderer, Collins and Tucker 1985; Rinderer, Tucker and Collins 

1982). Results of these investigations present quite a different 



140 
picture as to the factors leading to the success and resulting impact of 
the Africanized honey bee in South America. 

Factors Contribut ing to the Selective Advantage of 
Africanized Honey Bees in South America 

Colony Demographv 

Table 8-1 summarizes the factors affecting colony survival and 
reproductive success for both Africanized and European honey bees under 
tropical conditions in Venezuela. Studies comparing parameters of 
colony demography for Africanized and European honey bees under 
identical conditions in Venezuela have produced surprising results (see 
Chapters II-VI). These studies were based on the assumption that the 
life history of the Africanized honey bee population in South America 
(as well as the parental population in Africa) was characterized by a 
high reproductive rate. Demographic features that were expected to be 
correlated with this high rate of colony reproduction* or short swarm to 
swarm interval* were shorter worker bee development time, smaller worker 
bee size» more rapid queen development and maturation, increased egg 
laying and brood production, reduced brood mortality, and increased 
adult worker bee longevity. 

Colony demographic characteristics can be divided into two groups: 
those affecting the rate of colony growth and those affecting the time 
interval from swarming to the beginning of adult population increase. A 
rapid colony growth rate is most important for a high colony 
reproductive rate and is primarily a function of queen fecundity, adult 
worker longevity, and brood mortality (Brian 1965; Moeller 1961; Wilson 
1971). 



141 
Although Africanized queens have been reported to have greater egg 
laying rates than European queens (Fletcher 1978; Michener 1972, 1975; 
Ribbands 1953), under identical experimental conditions in Venezuela, 
there was no significant difference in queen fecundity during the 
initial colony growth period (Chapters V and VI). Also, European honey 
bee workers live longer (Winston and Katz 1981), giving European honey 
bee colonies a growth rate advantage with respect to this demographic 
parameter. 

Because of the relationship of worker longevity to colony growth 
rates, initial colony growth rates may be affected by the age structure 
of bees in a swann. Colonies established from swarms with older bees 
will have a more rapid decline in population, which will adversely 
affect colony growth because egg laying rates are a function of the 
number of bees in a colony (Moeller 1958). The age structure of 
Africanized swarms has been evaluated (Winston and Otis 1978) but there 
are no data for both Africanized and European swarms under similar 
conditions. 

Another parameter affecting colony growth rate is brood mortality, 
but there are no data available that simultaneously compare Africanized 
and European honey bees under identical conditions. Experimental and 
environmental conditions are particularly important with respect to this 
parameter. Rates of brood mortality can be as high as 50% and are 
affected by season, resource availability and colony adult population 
(Garofalo 1977; Merrill 1924; Woyke 1977). These high rates of brood 
mortality and/or brood cannibalism may function to regulate protein 
balance in honey bee colonies during protein (pollen) shortages (Weiss 
1984). Therefore, earlier studies comparing brood mortalities in 



142 
Africanized and European colonies (Winston, Dropkin and Taylor 1981), 
which were observed under very different conditions, need to be re- 
evaluated, and new studies should be undertaken. 

Also affecting colony growth rates is the extent to which brood 
rearing ceases during periods of resource dearth. As discussed earlier, 
Winston (1980b) reports that Africanized bees do not cease brood rearing 
to the extent observed for European bees and are therefore capable of 
rapid colony growth when conditions improve. European honey bees, under 
some tropical conditions, may have a sharp decline in brood rearing 
during resource shortages (Otis and Taylor 1979). However, these 
differences in brood rearing were not apparent when both Africanized and 
European honey bees were managed under identical conditions in Venezuela 
(Bolten, personal observation). Therefore, this behavior needs to be 
analyzed with both honey bee populations under a variety of tropical 
resource conditions to determine if there are differences, and whether 
the differences are a function of foraging behavior (see below) and/or 
intrinsic demographic parameters. 

The two most important factors affecting the interval from swarming 
to the beginning of adult population increase are worker development 
time and queen maturation (Chapters II and IV). Worker development time 
has previously been considered an important factor affecting the rate of 
colony growth (Fletcher 1977a, 1978; Fletcher and Tribe 1977a; Tribe and 
Fletcher 1977; Winston 1979b; Winston, Dropkin and Taylor 1981; Winston 
and Katz 1982; Winston, Taylor and Otis 1983). As discussed in Chapter 
II, this is incorrect and is probably a result of confusing models for 
colony growth (increase in the number of bees in the colony) with models 
for population growth (increase in the number of colonies). Population 



143 
growth models are designed for other species in which all individuals 
are potential reproductives. For honey bees, individual (or worker bee) 
development time is not equal to generation time. Worker bee 
development time only affects the interval between^a given change in egg 
laying rate and its resulting change in rate of adult emergence. The 
difference in worker development time between Africanized and European 
honey bees is only one day (Chapter II) and is trivial with respect to 
other factors affecting reproductive rates. 

Time from virgin queen emergence to initiation of oviposition also 
only affects the interval from swarming until adult population increase 
begins and not the rate of colony growth (Chapter IV). The results 
presented in Chapter IV for maturation rates of Africanized and European 
queens were unexpected. European honey bee queens began oviposition at 
an earlier age post-emergence than did Africanized queens. 
Reproduct ive Output 

Reproductive output is mainly determined by the swarm to swarm 
interval and the number of swarms produced per swarming cycle. The 
swarm to swarm interval is a function of all the colony demographic 
parameters discussed in the previous section and resource utilization 
parameters discussed in the next section. Although there are no data on 
the swarm to swarm intervals for Africanized and European honey bees 
under identical conditions, results from investigations of demographic 
parameters reported above suggest that if differences in swarm to swarm 
intervals exist, they would not be a result of demographic differences. 
Rather, if differences exist, they are hypothesized to be a result of 
differences in resource utilization between the two populations (see 
below) . 



144 
Although Africanized and European honey bees have not been compared 
under identical conditions, European colonies in North America (Kansas) 
produced the same number of small, secondary swarms, or afterswarms, per 
swarming cycle as did Africanized colonies in South America (French 
Guiana) (Otis 1980; Winston 1980a; Winston, Dropkin and Taylor 1981). 
These results are not directly comparable, but they do demonstrate that, 
at least under certain conditions, European honey bees can produce as 
many afterswarms as Africanized honey bees. Whether the number of 
afterswarms for European bees would be similar to Africanized bees under 
identical conditions needs to be analyzed. As discussed above, 
survivorship of small afterswarms would be much lower in temperate 
regions than in tropical regions, because of the energetic demands of 
temperate winters on honey bee colonies. 
Resource Utilization 

The most important factors leading to the success of Africanized 
honey bees in South America are associated with resource utilization: 
foraging behavior, brood production efficiency, worker bee size and 
absconding behavior. Although evaluated under different conditions, the 
foraging range of the parental African population is similar to that of 
European bees (Smith 1958b). However, under resource conditions typical 
of tropical regions, the foraging behavior of Africanized honey bees is 
significantly more successful than that of European honey bees 
(Rinderer, Bolten, Collins and Harbo 1984; Rinderer, Collins and Tucker 
1985). Their success is a result of more frequent solitary foraging and 
reduced recruitment when resources are dispersed and limited, as is 
characteristic of most tropical habitats (Rinderer, Bolten, Collins and 
Harbo 1984; Rinderer, Collins and Tucker 1985). Using artificial 



145 
flowers, Nunez (1973, 1979a, 1982) analyzed foraging behaviors of 
Africanized and European honey bees. Differences between the two 
populations were observed that were appropriate to having evolved under 
either temperate or tropical resource availability patterns. Hoarding 
cage studies also demonstrated differences in response between 
Africanized and European honey bees, suggesting differences in resource 
utilization behavior (Rinderer, Bolten, Harbo and Collins 1982). 

Increased numbers of bees in a colony are important for successful 
foraging, colony defense and reproduction (see Wilson 1971). When 
floral resources (nectar and pollen) are limited, a greater number of 
individual bees can be produced from a given amount of food if brood 
production is more efficient and/or if bees are smaller. There are no 
data available that were collected under identical conditions that allow 
comparison of brood production efficiency between Africanized and 
European honey bees, as measured by the ratio of developing brood to 
adult population for a range of different adult populations (Michener 
1964j Moeller 1961). However, their smaller size may result in 
increased brood production efficiency in Africanized bees, i.e., less 
food is necessary to produce smaller bees (Chapter III; Fletcher and 
Tribe 1977a; Tribe and Fletcher 1977). Therefore, with a limited food 
supply. Africanized honey bees could increase their population at a 
greater rate than European honey bees. With more successful foraging 
behavior and smaller bee size. Africanized colonies can grow rapidly 
under conditions where European colonies may not be able to survive. 

Another important difference between Africanized and European honey 
bees is the strategy used to survive during periods of food shortage. 
Honey bee colonies may either hoard sufficient quantities of food 



146 
(primarily honey) to sustain them during periods of resource dearth or 
the colonies can abscond (relocate or migrate) to another area where 
conditions may be better. Hoarding large surpluses of honey is 
characteristic of European honey bees in temperate regions. However, 
hoarding behavior may be disadvantageous in the tropics because colonies 
with large food surpluses may be more easily discovered by predators and 
less easily protected. Because predation has been a major evolutionary 
force for tropical honey bee populations (Seeley 1983; Seeley, Seeley 
and Akratanakul 1982), resource-induced absconding may be a better 
evolutionary alternative to hoarding. 

Resource- induced absconding occurs when an entire colony abandons a 
nest and is quite common in tropical species of Apis (A. florea , A. 
dorsata , and A. cerapa) and tropical populations of h, mellifera during 
periods of resource dearth (Winston, Otis and Taylor 1979; Winston, 
Taylor and Otis 1983; Woyke 1976). Africanized honey bees have a high 
rate of resource-induced absconding in South America (Winston, Otis and 
Taylor 1979). Resource- induced absconding behavior may not be 
advantageous in temperate regions because colonies may not have enough 
time once they settle in a new area to store sufficient food to 
successfully overwinter (Butler 1974). European honey bees generally do 
not abscond in either temperate or tropical regions (Butler 1974; 
Fletcher 1978; Winston, Otis and Taylor 1979; Winston, Taylor and Otis 
1983 ) . 

Whether resource- induced absconding behavior is a more successful 
strategy in the tropics than is hoarding behavior requires further study 
to determine the advantages and disadvantages of each strategy under 
tropical conditions. Both may be viable strategies and may not be 



147 
mutually exclusive (Winston, Otis and Taylor 1979). Fletcher suggests 
that resource- induced absconding may not always be appropriate, 
considering 

the distance that bees can fly in relation to the general 
distribution of their food plants. The maximum flight range 
is unlikely to exceed about 16 kilometres.. .and yet huge areas 
of Africa inhabited by honey-bees consist of more or less 
uniform grasslands and savannah. With certain exceptions, 
therefore, such as movements up and down mountain slopes and 
in and out of river valleys, there would appear to be little 
advantage in absconding in such areas, for within their flight 
range the bees would very often find only more of the same 
type of country they had left. (Fletcher 1975, p. 13). 

However, based on measurements of engorgement and estimates of metabolic 
rates, the maximum flight range of absconding colonies of honey bees has 
been calculated to be as great as 131 km (Otis, Winston and Taylor 
1981). In addition, periodic foraging while in-transit could extend the 
potential distance even further. Until comparative studies demonstrate 
the advantages of either resource- induced absconding or hoarding, 
resource- induced absconding behavior, which is common in tropical honey 
bee populations, is concluded to be an advantageous behavior under some 
tropical conditions. 
Predatio n and Colony Defense 

Ability to defend the nest from predators affects colony survival 
and therefore reproduction. The colony defense behavior characteristic 
of Africanized honey bees (Collins, Rinderer, Harbo and Bolten 1982; 
Stort 1974, 1975a, 1975b, 1975c, 1976) may be more effective than that 
of European honey bees. One aspect of colony defense behavior of 
Africanized honey bees, their stinging behavior, is so extreme that 
Africanized bees are a public health hazard for both humans and domestic 
animals (Taylor 1986). This colony defense behavior is particularly 
effective against vertebrate predators. Africanized honey bees in 



148 
Venezuela also reduce the size of their nest entrances to a greater 
extent than European honey bees, which helps to protect against 
invertebrate predators, particularly ants (Bolten, personal 
observation) . 

Besides colony defense, another response to predatlon Is to abscond 
(relocate). Because predatlon by both vertebrates and Invertebrates and 
infestation by wax moths ( Galleria mellonella and Achrola orlsella ) on 
honey bee colonies Is extensive In tropical regions, disturbance-Induced 
absconding would be an advantageous behavior and Is frequently observed 
In tropical honey bees (Fletcher 1976; Seeley 1983; Seeley, Seeley and 
Akratanakul 1982; Winston, Taylor and Otis 1983). Disturbance-Induced 
absconding was more frequently observed In Africanized honey bees than 
In European honey bees under similar conditions In Venezuela-, 
particularly with respect to attacks by ants (Bolten, personal 
observation) . 

In addition to predatlon on the colony-level. Africanized worker 
bees may have behaviors that are better adapted to avoiding predators 
and parasites while foraging. The rapid, zig-zag flight of worker bees 
In the African parental population may be more advantageous In avoiding 
predators (Invertebrate as well as vertebrate) compared with the slower, 
less erratic flight of European honey bees (Fletcher 1977b). Also, 
queen honey bees on mating flights are susceptible to predators. 
Fletcher (1977b) suggests that queens from the African parental 
population have shorter mating flights than European queens which may 
reduce predatlon. The differences In flight patterns and behaviors of 
Africanized and European honey bees need to be Investigated under 
similar conditions. 



149 
Nest Sit es and Cavity Volume 

Fletcher (1976) suggested another adaptive advantage that 
Africanized bees have is their ability to utilize a greater variety of 
nest sites. Fletcher may be confusing cause with effect when he 
suggests that it is this ability that "has enabled them to establish 
themselves in areas not previously inhabited by honey-bees at all" 
(1976, p. 6). A more likely explanation for the success of Africanized 
bees in those areas would be their ability to utilize the particular 
nectar and pollen resources available. That is, without a more 
efficient utilization of resources. Africanized honey bees would not be 
able to exploit these other habitats irrespective of their ability to 
utilize a greater variety of nest sites. As discussed above. 
Africanized honey bees are more successful foragers than are European 
honey bees under resource conditions typical of tropical habitats. 

Nest cavity volume is another factor affecting reproduction in 
honey bees. One of the stimuli for reproductive swarming is brood-nest 
crowding (Baird and Seeley 1983; Simpson 1966, 1973; Simpson and Riedel 
1963; Winston and Taylor 1980). Colonies inhabiting smaller cavities 
become crowded more rapidly and have a higher tendency to swarm. 
Colonies established in large cavities would be less crowded and have a 
lower rate of swarming. In temperate regions, small cavities would be 
selected against because there would be less volume available for 
storing surplus honey to enable the colony to overwinter. Therefore, 
Seeley proposed that nest-cavity volume may "regulate mature colony size 
at an optimum between small colonies with low survivorship and large 
colonies with low fertility" (Seeley 1977, p. 226). Jaycox and Pa rise 
(1980, 1981) found that honey bees from northern Europe selected larger 



150 
nest cavities than did honey bees from southern Europe. Southern 
European winters would be much less severe than those in northern 
Europe, and therefore the need for larger nest cavities to store large 
food surpluses is less important. 

For tropical honey bee populations, there may be a selective 
advantage for smaller nest cavity volumes, e.g., to facilitate 
protection against infestation from wax moths ( Galleria mellonella and 
^cliroja grisella) (Fletcher 1976). Africanized honey bees utilize a 
wider variety of nest sites than do European honey bees, including 
smaller cavity volumes (Fletcher 1976). The negative factors associated 
with smaller nest cavities may be absent in the tropics because there is 
less need to store large surpluses to survive periods of resource 
dearth— periods are generally shorter and less costly with respect to 
energetic demands for maintaining proper brood nest temperature. In 
addition, honey bees that evolved in the tropics commonly abscond during 
periods of resource dearth as opposed to hoarding surplus food. 

Although nest cavity choice for Africanized and European honey bees 
has not been studied under identical conditions, nest cavities selected 
by Africanized honey bees in Venezuela were not smaller than cavities 
selected by European honey bees in temperate regions (Rinderer, Collins, 
Bolten and Harbo 1981; Rinderer, Tucker and Collins 1982). Because of 
the importance of nest cavity volume to reproductive rates, nest cavity 
volume for both populations needs to be investigated under identical 
conditions. 
Density-Dependent Factors Ren ulatinc;] Queen Rearing 

Other parameters that might account for differences in reproductive 
rates between Africanized and European honey bees may be certain 



151 
density-dependent factors that are responsible for regulating queen 
rearing in colonies preparing to reproduce. In European honey bees, 
initiation of queen rearing prior to reproduction is not a result of a 
decrease in queen pheromone production (Seeley and Fell 1981). Two 
other possibilities are suggested by Seeley and Fell (1981). First, 
there may be failure to adequately disperse queen pheromone in crowded 
colonies prior to swarming. And, second, worker bee response to queen 
pheromone may change prior to swarming. 

Threshold levels for queen pheromone that inhibit queen rearing in 
worker bees may be different for Africanized and European honey bees. 
Also, dispersal of queen pheromone by "messenger" bees (Seeley 1979) may 
be different for Africanized bees compared with European bees. Baird 
and Seeley (1983) developed an equilibrium model to explain the 
regulation of queen rearing in colonies preparing to reproduce. Their 
model postulated that "there is a balance between nurse bees becoming 
inhibited from queen rearing and nurses losing their inhibition, and 
that whether a colony does or does not rear queens reflects the 
equilibrium percentage of inhibited nurses" (Baird and Seeley 1983, p. 
221). Therefore, differences between Africanized and European honey 
bees with respect to density-dependent factors regulating queen rearing 
may result in differences in reproductive rates by affecting: 1) adult 
population size when colonies reproduce; 2) prime swarm size; and 3) 
number of afterswarms. Some of these density-dependent factors have 
been compared for Africanized bees in South America with European bees 
in North America under different environmental and experimental 
conditions (Winston, Dropkin and Taylor 1981). Unfortunately, there are 
no data collected under identical conditions that allow valid 



152 
comparisons to be made between Africanized and European honey bees that 
enable any density-dependent factors responsible for the reproductive 
rates and success of Africanized honey bees in South America to be 
identified. 
Conclusion 

In tropical regions, the success of Africanized honey bees compared 
with European honey bees is not a function of any intrinsic differences 
in colony demography. Rather, it must be concluded that the success of 
Africanized honey bees is due primarily to their ability to efficiently 
utilize tropical resources, enabling them to survive and reproduce under 
conditions where European honey bees are frequently not able to survive. 
If European honey bee colonies are not able to survive and/or grow under 
some of the tropical resource conditions of South America, they 
obviously cannot reproduce. It is precisely because the European honey 
bees were not successful foragers (= honey producers) in most tropical 
regions of Brazil that African honey bees were imported into Brazil 
(Goncalves 1974, 1975, 1982; Woyke 1969). 

The efficient utilization of tropical resources by Africanized 
honey bees is a result of a set of adaptive behaviors involving solitary 
foraging, reduced recruitment, increased brood production efficiency 
because of smaller worker bee size, and both resource- induced and 
disturbance-induced absconding. These characteristics, combined with an 
effective colony defense behavior, give Africanized honey bees a 
selective advantage that results in increased survivorship, increased 
colony growth rates and ultimately increased reproduction, which is 
responsible for their rapid dispersal and high population densities. 



153 
The Africanized honey bees studied in Venezuela are only a small 
sample of the total Africanized honey bee population 1n South and 
Central America and represent only a fraction of the variation within 
the population, particularly if we consider that Africanized honey bees 
are a result of hybridization. Nevertheless, the results presented in 
the foregoing chapters demonstrate that at least some portion of the 
Africanized honey bee population is similar to the European honey bee 
population with respect to the demographic parameters analyzed. 

Potential Impact of Africanized Honey Bees in North America 
The selective advantage of Africanized honey bees in South America 
will be lost as they disperse north into temperate regions. European 
honey bees will have the selective advantage in temperate regions 
because of their particular behavioral repertoire which is better 
adapted to temperate conditions. However, because the populations can 
interbreed successfully, negative characteristics of the Africanized 
population, e.g., their stinging behavior, may become genetically 
introgressed into the European population of North America and therefore 
widespread wherever honey bees can survive (Chapter VII). A more 
optimistic scenario is that the large population of European honey bees 
in Mexico will slow the spread of African genes because of competition 
for available resources as well as through hybridization. Therefore, 
through selection, hybridization, and competition, the impact of 
Africanized honey bees may be minimized in North America (Chapter VII). 



154 



TABLE 8-1. Factors affecting colony survival and reproductive success 
for Africanized and European honey bees In Venezuela. 



FACTOR 



Brood Production 
Efficiency 

Bee Size 

Resource-Induced 
Absconding 

PREDATION 

Colony Defense 

Disturbance-Induced 
Absconding 

Flight Behavior 

NEST CAVITY VOLUME 

DENSITY-DEPENDFNfT FACTOR.^ 



POPULATION 
WITH ADVANTAGE 



COLONY DEMOGRAPHY 

Growth Rate 

Egg Laying Rate 

Worker Longevity 

Swarm Age Structure 

Brood Mortal ity 

Brood Production 
during Resource Dearth 

Interval^ 

Worker Development Time 

Queen Maturation 

REPRODUCTIVE OUTPUT 

Number of Afterswarms 
per Swarming Cycle 

RESOURCE UTILIZATION 

Foraging Behavior 



No Difference 
European 
No Data 
No Data 
No Data 



Africanized 
European 



No Data 



Africanized 



No Data 

Africanized 
Africanized 



Africanized 

Africanized 

No Data 
No Data 
No Data 



REFERENCES 



Chapter VI; Chapter V 
Winston & Katz 1981 



Chapter II 
Chapter IV 



Rinderer, Bolten, Collins 
& Harbo 1984; Rinderer, 
Bolten, Harbo & Collins 
1982; Rinderer, Collins & 
Tucker 1985; Nunez 1979, 
1982; Winston & Katz 1982 



Chapter III 

Winston, Otis & Taylor 1979 
Winston, Taylor & Otis 1983 



Collins, Rinderer, Harbo & 
Bolten 1982 

Bolten, pers. observation; 
Winston, Taylor & Otis 1983 



^Interval from swarming to beginning of population increase. 



APPENDIX A 

WORKER BEE DEVELOPMENT TIMES AND MORTALITY 

DURING DEVELOPMENT 



TABLE A-1. Comparison of worker bee development time (in days) for 

Africanized and European honey bees: median, (range), mean 
+ SD, (n = sample size). All development measured in 
European comb cell size with European nurse bees. 



1 EGG 

GENOTYPES 

1 


UNSEALED BRCJOD 


SEALED BROOD 


TOTAL 
DEVELOPMEMT^ 


1 

I AFRICANIZED 








A53 
(n = 25) 

1 


5.0 

(4-5) 

4.8 +0.4 


11.0 
(11-12) 
11.4 +0.5 


19.0 

(19-20) 

19.2 + 0.4 


1 A26 

i (n = 9) 

1 
1 


5.0 
(5) 
5.0 + 


11.0 
(11-12) 
11.3 +0.5 


19.0 
(19-20) 
19.3 +0.5 


A25 
(n = 19) 

1 


5.0 

(4-5) 

4.6 +0.5 


11.0 
(11-12) 
11.4 +0.5 


19.0 
(19) 
19.0 +0 


COMBINED 
(n = 53) 


5.0 

(4-5) 

4.8 +0.4 


11.0 
(11-12) 
11.4 +0.5 


19.0 

(19-20) 

19.2 + 0.4 


EUROPEAN 








W18 
(n = 28) 


5.0 

(4-5) 

4.8 +0.4 


12.0 
(11-13) 
12.0 + 0.3 


20.0 
(19-20) 
19.8 +0.4 


HI 
(n = 19) 


5,0 

(5-6) 

5.4 +0.5 


12.0 

(11-13) 

12.1 +0.4 


20.0 
(20-21) 
20.5 +0.5 


Y(A5) 
(n = 26) 


5.0 

(4-5) 

4.8 + 0.4 


12.0 

(11-12) 

12.0 +0.2 


20.0 

(19-20) 

19.8 +0.4 


COMBINED 
(n = 73) 


5.0 

(4-6) 

5.0 +0.5 


12.0 
(11-13) 
12.0 +0.3 


20.0 
(19-21) 
20.0 + 0.5 


analyses'^ 


NS 


P<0.001 


P<0.001 



Total development = time from oviposit ion to adult emergence. 
Kolmogorov-Smirnov one-tailed test, chi-square distribution, df 
alpha = 0.05 (Siegel 1956). Combined samples used for analyses, 

156 



= 2, 



157 

TABLE A-2. Mortality during different developmental stages. Mortality 
was measured in European comb cell size with European nurse 
bees. 





GENOTYPES 


^1^ 


^2 


4^ 


1 d 
'-2 


SB® 


N^^ 


AFRICANIZED EGG 














A53 










5 








30 


A26 










— 


199 





28 


A25 




1 





9 








29 


EUROPEAN EGG GENOTYPES 














W18 













2 





30 


HI 










3 


8 





30 


Y(A5) 










4 








30 



portal ity during first 24 hours in test colony (acceptance). 

^Mortality between 24-72 hours (before hatching). 

Sj^ortality between 72-96 hours (at time of hatching). 

2^ortality during older larval stages, before sealing. 

portal ity during the pupal stage. 

^N = total eggs monitored. 

9Not distinguished between l-^ and L2. 



APPENDIX B 
HONEY BEE SIZE, COMB CELL SIZE AND SIZE VARIATION 



TABLE B-1. Coefficients of variation (CV) of worker bees in honey 
bee and bumble bee colonies calculated from data 
presented in the references. 



CV 



REFERENCES 



Honey bees ( Apis mell if era ) 
Weights 

Adult (fresh) 



Adult (dry) 



Linear Measurements 
Length forewing 
Width forewing 
Length proboscis 

Bumble bees ( Bombus ) 
Weights 

Adult (fresh) 
(Bombgs aqrorum) 

Linear Measurements'^ 

Length radial cell 
(BombM$ fervidus) 



0.4-0.6 
4.0-4.5^ 
5.4-7.0"^ 
10.7-11.2^ 

4.8-8.1 

4.1-4.5^ 

4.1-6.7^ 



1.5-1.6 
2.2-2.5 
1.6-1.9 



31.0-36.7 



7.4-13.8 



Abdellatif 1965 
Bolten (Table B-2) 
Bolten (Table B-2) 
Kerr and Hebling 1964 

Grout 1937 

Bolten (Table B-2) 

Bolten (Table B-2) 



Grout 1937 
Grout 1937 
Grout 1937 



Brian 1952 



Medler 1965 



^European genotypes. 

Africanized genotypes in South America. 

CV may be high as a result of variable engorgement during 4 hour delay 
from emergence to weighing. 

Different linear measurements for Bombus are significantly correlated 
(P<0.01, Medler 1962). 



159 



160 



TABLE B-2. Africanized and European adult honey bee weights (mg): 

mean + SD, coefficient of variation (CV), (sample size). 



GENOTYPE 



COMB CELL 
TYPE^ 



FRESHLY 
B^ERGED 



DRIED^ 



CORRELATIONS'^ 



AFRICANIZED 
A26 

AST 

B39 

A60 

EUROPEAN 
WE2 

Y(K) 



EUR 



EUR 



EUR 



AFR 



EUR 



EUR 



94.9 +5.1 
CV = 5.4 
(28) 

88.6 + 6.2 
CV = 7.0 
(30) 

87.4 + 4.7 
CV = 5.4 
(16) 

95.2 +5.1 
CV = 5.4 
(30) 



107.1 +4.8 
CV = 4.5 
(29) 

116.1 +4.7 
CV = 4.0 
(27) 



13.5 + 0.6 
CV = 4.4 
(28) 

11.9 + 0.8 
CV = 6.7 
(30) 

12.3 + 0.5 
CV = 4.1 
(16) 

13.2 + 0.6 
CV = 4.5 
(30) 



14.6 + 0.6 
CV = 4.1 
(29) 

15.5 +0.7 
CV = 4.5 
(27) 



EUR = European comb cell diameter = 5.4 mm. 

AFR = Africanized comb cell diameter = 4.8 mm. 
■^Dried at 50°C for 48 hrs. 
^Pearson's correlation coefficient, alpha = 0.05; 

** = P<0.01; *** = P<0.001. 



*** 



»** 



NS 



*** 



** 



*** 



161 



TABLE B-3, 



Comparison of Africanized and European comb cell diameter 
and comb cell volume: mean + SD, range, CV, (sample size). 



AFRICANIZED COMB CELLS 



DIAMETER^ 
(mm) 



4.8 +0.1 
4.6 - 4.9 
CV = 2.1 
(50) 



volume'' 

(ml X 10~3 



184.6 + 15.8 
160 - 215 
CV = 8.6 
(50) 



CORRELATIONS^ 



NS 



EUROPEAN COMB CELLS 



ANALYSES® 



5.4 +0.05 
5.4 - 5.5 
CV = 1.0 
(30) 



P<0.001 



264.3 +23.5 
225 - 300 
CV = 8.9 
(30) 



P<0.001 



NS^ 



^Determined by measuring 10 horizontal, consecutive cells; cell-wall 

thickness not considered. 

Cell volume determined by filling cells with water with a pipette. 

Spearman's rank correlation coefficient, alpha = 0.05. 
"Negative correlation, P<0.01. 
®t-test, one-tailed. 



162 



TABLE B-4. Comparison of Africanized and European comb cell diameter 
and comb cell depth: mean + SD, range, (sample size). 

DIAMETER^ DEPTH 

(mm) (mm) CORRELATIONS'^ 



AFRICANIZED COMB CELLS 4.8+0.1 11.8+0.2 NS 

4.7 - 4.9 11.4 - 12.3 
(17) (17) 



EUROPEAN COMB CELLS 5.4 +0.05 12.2 +0.4 NS 

5.3 - 5.4 11.5 - 12.8 
(26) (26) 



ANALYSES*^ P<0.001 P<0.001 



Determined by measuring 10 horizontal, consecutive cells; cell-wall 

thickness not considered. 

■^Spearman's rank correlation coefficient, alpha = 0.05. 
^t-test, one-tailed. 



163 



TABLE B-5. Changes in European worker bee pupal weight (mg) with 
changes in pupal age: mean + SD, (sample size). 
Fresh weights were measured in Gainesville, Florida. 



AGE (DAYS POST-OVIPOSITION) 



11.5 


12.5 13.5 


14.5 


15.5 


16.5 


17.5 


145.8 


145.5 141.9 


140.9 


141.1 


139.1 


138.0 


+ 4.8 


+ 3.8 +3.8 


+ 4.2 


+ 4.6 


+ 4.0 


+ 4.8 


(10) 


(45) (31) 


(38) 


(31) 


(32) 


(34) 


A 


B C 


D 


E 


F 


G 


ANALYSES 


CDEF 

A X B 
B X C 
C X D 
D X E 
E X F 
F X G 


NS^ 

ns^ 

NS 
NS 
NS 
NS 
NS 









^One-way analysis of variance, alpha = 0.05. 
°t-test, two-tailed, alpha = 0.05. 



APPENDIX C 
CHANGES IN QUEEN PUPAL WEIGHT WITH AGE 



TABLE C-1. Changes 1n European queen pupal weights (mg) with 

changes in age: mean + SD, (sample size). Queen pupal 
weights were measured in Baton Rouge, Louisiana. 



DAYS POST-OVIPOSITION 



9 


10 


11 


12 


13 


14 


311.6 
+ 12.4 
(10) 

A 


291.6 
+ 8.3 
(10) 

B 


294,6 
+ 9.7 
(10) 

C 


293.4 
+ 8.4 
(10) 

D 


287.4 
+ 11.2 
(10) 

E 


274.1 
+ 16.9 
(5) 

F 


ANALYSES^ 




ABCDEF 
BCDE 


P<0.001 
NS 







^One-way analysis of variance, alpha = 0.05. 



165 



APPENDIX D 
ACCURACY OF TECHNIQUE USED TO ESTIMATE NUMBERS OF BEES IN A COLONY 



TABLE D-1. Accuracy of technique used to estimate number of bees in 
colony. 



ESTIMATED^ DIFFERENCE % DIFFERENCE 



SAMPLE 


NO. 


COUNTED 


1 




2284 


2 




2906 


3 




3085 


4 




3079 


5 




2657 


e 




2046 



2324 40 1.8 

2918 12 0.4 

3080 5 0.2 

3124 45 1.5 

2717 60 2.2 

2102 56 2.7 



a 



■The number of adult bees was estimated by determining the mean 
individual bee weight from three, 150-200 bee samples. The total 
weight of bees in each sample was then divided by the mean individual 
bee weight to get an estimate of the total number of bees in each 
sample. 



167 



LITERATURE CITED 



Abdellatif, M.A. 1965. Comb cell size and Its effect on the body 

weight of the worker bee. Apis melllfera L. Am. Bee J. 105:86-87, 

Adams, J,, E.D. Rothman, W.E. Kerr and Z.L. Paulino. 1977. Estimation 
of the number of sex alleles and queen matlngs from diploid male 
frequencies In a population of Apis melHfera . Genetics 86: 
583-596. 

Alles, P. 1961. Role of queen and worker bees In the hereditary 
transmission of certain characteristics. Aplc. Absts. 12:167. 

Alpatov, W.W. 1929. Blometrlcal studies on variation and races of the 
honey bee. Quart. Rev. Biol. 4:1-58. 

Alpatov, W.W. 1933. South African bees blometrlcally Investigated. 
Bee World 14:62-64. 

Anderson, R.H., B. Buys and M.F. Johannsmeler. No date. Beekeeping In 
South Africa. Technical Services Bulletin 394. South African 
Dept. of Agriculture, Pretoria. 

Balrd, D.H., and T.D. Seeley. 1983. An equilibrium theory of queen 
production 1n honeybee colonies preparing to swarm. Behav. Ecol. 
Sodoblol. 13:221-228. 

Baudoux, U. 1933. The Influence of cell size. Bee World 14:37-41. 

Beetsma, J. 1979. The process of queen-worker differentiation In the 
honeybee. Bee World 60:24-39. 

Boch, R. 1957. Rassenmablge Unterschlede bel den Tanzen der 
Honlgblene. Z, Vergl. Physiol. 40:289-320. 

Boch, R., and C.A. Jamleson. 1960. Relation of body weight to 
fecundity In queen honeybees. Can. Ent, 92:700-701. 

Boch, R., D.A. Shearer and J.C. Young. 1975. Honeybee pheromones: 
Field tests of natural and artificial queen substance. J. Chem. 
Ecol. 1:133-148. 

Bodenhelmer, F.S. 1937. Studies In animal populations, II. Seasonal 
population trends of the honey-bee. Quart. Rev. Biol. 12:406-425. 



168 



169 

Bodenhelmer, F.S., and A. Ben-Nerya. 1937. One-year studtes on the 
biology of the honey-bee in Palestine. Ann. Appl. Biol. 24: 
385-403. 

Bolten, A.B., and J.R. Harbo. 1982. Numbers of spermatozoa In the 

spermatheca of the queen honeybee after multiple Inseminations with 
small volumes of semen. J. Aplc. Res. 21:7-10. 

Brian, A.D. 1952. Division of labour and foraging In Bomb us aororum 
Fabrldus. J. Anim. Ecol. 21:223-240. ^ 

Brian, M.V. 1965. Social Insect populations. Academic Press, London. 

Brother Adam. 1966. In search of the best strains of bees. 
Ehrenwlrth, Munich. 

Buchmann, S.L. 1982. Africanized bees confirmed in Panama. Am. Bee J. 
122:322. 

Buchner, R. 1955. Effect on the size of workers of restricted space 
and nutrition during larval development. Apic. Absts. 6:15. 

Butler, C.G. 1971. The mating behaviour of the honeybee ( Ad is 
fflSllifsna L.). J. Ent. (A) 46:1-11. '^ 

Butler, C.G. 1974. The world of the honeybee. Collins, London. 

Butler, C.G., D.H. Calam and R.K. Callow. 1967. Attraction of Apis 
m$11lf?ra drones by the odours of the queens of two other species 
of honeybees. Nature 213:423-424. 

Cantwell, G.E. 1974. The African (Brazilian) bee problem. Am. Bee J. 
114:368-372. 

Carlson, D.A., and A.B. Bolten. 1984. Identification of Africanized 
and European honey bees, using extracted hydrocarbons. Bull. Ent. 
Soc. Am. 30:32-35. 

Collins, A.M., T.E. Rinderer, J.R. Harbo and A.B. Bolten. 1982. Colony 
defense by Africanized and European honey bees. Science 218:72-74. 

Crewe, R.M., and H. Hastings. 1976. Production of pheromones by 

workers of Ajils. melllfepa adansonii . J. Apic. Res. 15:149-154. 

Daly, H.V., and S.S. Balling. 1978. Identification of Africanized 
honeybees in the Western Hemisphere by discriminant analysis. J. 
Kans. Ent. Soc. 51:857-869. 

Daly, H.V., K. Hoelmer, P. Norman and T. Allen. 1982. Computer- 
assisted measurement and identification of honey bees (Hymenoptera: 
Apidae). Ann. Ent. Soc. Am. 75:591-594. 

Darwin, C. 1958. The origin of species. The New American Library, 
Inc., New York. (Original work published 1859). 



170 

Dietz, A. 1978. An anatomical character suitable for separating drone 
honey bees of Ap1s mell If era 1 1 gust lea from Apis mell 1fera 
adansonll . Pages 102-106 In Aplcultura em cl Ima quente. 
Ap1mond1a» Florlanopol is, Brazil. 

Dietz, A., R. Krell and F.A. Elschen. 1985. Preliminary investigation 
on the distribution of Africanized honey bees in Argentina. 
Apidologie 16:99-108. 

DuPraw, E.J. 1965. Non-Linnean taxonomy and the systematics of honey 
bees. Syst. Zool. 14:1-24. 

Eckert, J.E. 1934. Studies on the number of ovarioles in queen 

honeybees in relation to body size. J. Econ. Ent. 27:629-635. 

Ewel, J. J., and A. Madriz. 1968. Zonas de vida de Venezuela. 
Ministerio de Agricultura y Cria, Caracas. 

Fell, R.D., J.T. Ambrose, D.M. Burgett, D. De Jong, R.A. Morse and T.D. 
Seeley. 1977. The seasonal cycle of swarming in honeybees. J. 
Apic. Res. 16:170-173. 

Fletcher, D.J.C. 1975. New perspectives in the causes of absconding in 
the African bee ( Apis mell if era adansonii L.), Part I. S. Afr. Bee 
J. 47:11-14. 

Fletcher, D.J.C. 1976. New perspectives in the causes of absconding In 
the African bee ( Apis mell if era adansonll L.), Part II. S. Afr. 
Bee J. 48:6-9. 

Fletcher, D.J.C. 1977a. A preliminary analysis of rapid colony 

development in Apis mell if era adansonii . Pages 144-145 in Proc. 
Eighth Int. Congress, Int. Union for Study of Social Insects. 
Wageningen, Netherlands. 

Fletcher, D.J.C. 1977b. Evaluation of introductions of European honey- 
bees into southern and eastern Africa. Pages 146-147 in Proc. 
Eighth Int. Congress, Int. Union for Study of Social Insects. 
Wageningen, Netherlands. 

Fletcher, D.J.C. 1978. The African bee. Apis mell if era adansonii , in 
Africa. Ann. Rev. Ent. 23:151-171. 

Fletcher, D.J.C, and G.D. Tribe. 1977a. Swarming potential of the 
African bee. Apis mell if era adansonii L. Pages 25-34 in D.J.C. 
Fletcher (ed,), African bees: Taxonomy, biology and economic use. 
Apimondia, Pretoria. 

Fletcher, D.J.C, and G.D. Tribe. 1977b. Natural emergency queen 

rearing by the African bee &.. UL- adansonii and its relevance for 
successful queen production by beekeepers, II. Pages 161-168 in 
D.J.C. Fletcher (ed.), African bees: Taxonomy, biology and 
economic use. Apimondia, Pretoria. 



171 

Fletcher, D.J.C.» and G.D. Tribe. 1977c. Natural emergency queen 

rearing by the African bee A. a. adanson-fi and its relevance for 
successful queen production by beekeepers, I. Pages 132-140 in 
D.J.C. Fletcher (ed.), African bees: Taxonomy, biology and 
economic use. Apimondia, Pretoria. 

Free, J.B. 1965. The allocation of duties among worker honeybees. 
Symp. Zool. Soc. (London) 14:39-59. 

Fyg, W. 1959. Normal and abnormal development in the honeybee. Bee 
World 40:57-66, 85-96. 

Garofalo, C.A. 1977. Brood viability in normal colonies of Apis 
mellifera . J. Apic. Res. 16:3-13. 

Gary, N.E. 1975. Activities and behavior of honey bees. Pages 185-264 
in Dadant and Sons Ceds.), The hive and the honey bee. Dadant and 
Sons, Inc., Hamilton, Illinois. 

Glushkov, N.M. 1958. Experiments on combs with enlarged cells. Apic. 
Absts. 9:102. 

Goncalves, L.S. 1974. Comments on the aggressiveness of the 
Africanized bees in Brazil. Am. Bee J. 114:448-450. 

Goncalves, L.S. 1975. Do the Africanized bees of Brazil only sting? 
Am. Bee J. 115:8-10. 

Goncalves, L.S. 1982. The economic impact of the Africanized honey bee 

in South America. Pages 134-137 in M.D. Breed, CD. Michener and 

H.E. Evans (eds.). The biology of social insects. Westview Press, 
Boulder, Colorado. 

Gould, J.L. 1982. Why do honey bees have dialects? Behav. Ecol. 
Sociobiol. 10:53-56. 

Grout, R.A. 1937. The influence of size of brood cell upon the size 
and variability of the honeybee. Research Bulletin 218, 
Agricultural Experiment Station, Iowa State College of Agriculture 
and Mechanic Arts, Ames, 

Hall, H.G. In press. DNA differences found between Africanized and 
European honeybees. Proc. Natl. Acad. Sci., USA. 

Harbo, J.R. 1979. Storage of honeybee spermatozoa at -196°C. J. Apic. 
Res. 18:57-63. 

Harbo, J.R., and A.B. Bolten. 1981. Development times of male and 
female eggs of the honey bee. Ann. Ent. Soc. Am. 74:504-506. 

Harbo, J.R., A.B. Bolten, T.E. Rinderer and A.M. Collins. 1981. 

Development periods for eggs of Africanized and European honeybees. 
J. Apic. Res. 20:156-159. 



172 

Harbo, J.R., and T.I. Szabo. 1984. A comparison of Instrumentally 
Inseminated and naturally mated queens. J. Ap1c. Res. 23:31-36. 

Harp, E.R. 1973. A specialized system for multiple rearing of quality 
honeybee queens. Am. Bee J. 113:256-258, 261. 

Helnrlch, B. 1979a. Bumblebee economics. Harvard Un1v. Press, 
Cambridge, Mass. 

Helnrlch, B. 1979b. Thermoregulation of African and European honeybees 
during foraging, attack, and hive exits and returns. J. Exp. Biol. 
80:217-229. 

Holdrldge, L.R. 1964. Life zone ecology. Tropical Science Center, San 
Jose, Costa Rica. 

Hooplngarner, R., and C.L. Farrar. 1959. Genetic control of size of 
queen honey bees. J. Econ. Ent. 52:547-548. 

Jay, S.C. 1963. The development of honeybees In their cells. J. Aplc. 
Res. 2:117-134. 

Jaycox, E.R., and S.G. Parlse. 1980. Homes Ite selection by Italian 
honey bee swarms, Apjs melllfera Hyustlca (Hymenoptera: Apldae). 
J. Kans. Ent. Soc. 53:171-178. 

Jaycox, E.R., and S.G. Parlse. 1981. Homes Ite selection by swarms of 
black-bodied honey bees. Apt? melllfera caucasica and A. eq.. carnjca 
(Hymenoptera: Apldae). J. Kans. Ent. Soc. 54:697-703. 

Jeffree, E.P. 1955. Observations on the decline and growth of honey 
bee colonies. J. Econ. Ent. 48:723-726. 

Johansson, T.S.K., and M.P. Johansson. 1973. Methods for rearing 
queens. Bee World 54:149-175. 

Kerr, W.E. 1967. The history of the introduction of African bees to 
Brazil. S. Afr. Bee J. 39:3-5. 

Kerr, W.E., and D. Bueno. 1970. Natural crossing between Apis 
mglllf^ra adan^onH and Ap1s men if era liqustica . Evolution 
24:145-155. 

Kerr, W.E., L.S. Goncalves, L.F. Blotta and H,B. Maciel. 1972. 
Biologia comprada entre as abelhas italianas ( Apis melllfera 
1igu?tica) Afrlcana (Apis melllfera adansonli ) e suas hibridas. 
Pages 151-185 in Anais do 1° Congresso Brasileiro de Apicultura. 
Congresso Brasileiro de Apicultura, Florianopol is, Brazil. 

Kerr, W.E., and N.J. Hebling. 1964. Influence of the weight of worker 
bees on division of labor. Evolution 18:267-270. 

Koeniger, N., and H.N. P. Wijayagunasekera. 1976. Time of drone flight 
in the three Asiatic honeybee species. J. Apic. Res. 15:67-71. 



173 

Krell, R., A. Dietz and F.A. Eischen. 1985. A preliminary study on 

winter survival of Africanized and European honey bees in Cordoba* 
Argentina. Apidologie 16:109-118. 

Kulzhinskaya* K.P, 1956. The role of the food factor in the growth of 
bees. Apic. Absts. 7:177. 

Laidlawj H.H.» Jr. 1979. Contemporary queen rearing. Dadant and Sons, 
Inc., Hamilton, Illinois. 

Laidlaw, H.H., Jr., and J.E. Eckert. 1962. Queen rearing. Univ. of 
California Press, Berkeley. 

Lindauer, M. 1953. Division of labour in the honeybee colony. Bee 
World 34:63-73, 85-90. 

Mackensen, 0. 1964. Relation of semen volume to success in artificial 
inseminations of queen honey bees. J. Econ. Ent. 57:581-583. 

Mackensen, C, and W.C. Roberts. 1948. A manual for the artificial 
insemination of queen bees. USDA-ARA Bur. Ent. and Plant Quar. 
ET-250. 

Mackensen, 0., and K. Tucker. 1970. Instrumental insemination of queen 
bees. Agric. Hdbk. 390. USDA, Washington, D.C. 

McDowell, R. 1984. The Africanized bee in the United States: What 
will happen to the U.S. beekeeping industry? Agric. Econ. Rpt. 
519. USDA, Washington, D.C. 

McGregor, S.E. 1938. Environmental factors and size variations in 
honeybee appendages. J. Econ. Ent. 31:570-573. 

McLellan, A.R. 1978. Growth and decline of honeybee colonies and 
inter-relationships of adult bees, brood, honey and pollen. J. 
Appl. Ecol. 15:155-161. 

Medler, J.T. 1962. Morphometric studies in bumble bees. Ann. Ent. 
Soc. Am. 55:212-218. 

Medler, J.T. 1965. Variation in size in the worker caste of Bombus 
fervidus (Fab.). Pages 388-389 in P. Freeman (ed.), Proc. Xllth 
International Congress of Entomology. Royal Entomological Society 
of London, London. 

Melampy, R.M., and E.R. Willis. 1939. Respiratory metabolism during 

larval and pupal development of the female honeybee ( Apis mellifera 
L.). Physiol. Zool. 13:283-293. 

Mel'nichenko, A.N. 1962, Experiments in directed alteration of the 

characteristics of female and male honeybees in rearing in the hive 
of a different variety. Biol. Abs. 39:661. 



174 

Merrill, J.H. 1924. Observations on brood-rearing. Am. Bee J. 64: 
337-338. 

Michener, CD. 1964. Reproductive efficiency in relation to colony 
size in hymenopterous societies. Insectes Sociaux 11:317-341. 

Michener, CD. 1972. Final report of the Committee on the African 
Honey Bee. Nat. Res. Counc, Nat. Acad. Sci,, Washington, D.C 

Michener, CD. 1974. The social behavior of the bees. Belknap Press, 
Cambridge, Mass. 

Michener, CD. 1975. The Brazilian bee problem. Ann. Rev. Ent. 
20:399-416. 

Michener, CD. 1979. Biogeography of the bees. Ann. Missouri Bot. 
Gard. 66:277-347. 

Milum, V.G. 1930. Variations in time of development of the honeybee. 
J. Econ. Ent. 23:441-447. 

Moeller, F.E. 1958. Relation between egg-laying capacity of queen bee 
and populations and honey production of their colonies. Am. Bee J. 
98:401-402. 

Moeller, F.E. 1961. The relationship between colony populations and 
honey production as affected by honey bee stock lines. Production 
Research Report 55. USDA, Washington, D.C 

Morse, R.A. 1984. The mating behavior of African queens. Glean. Bee 
Cult. 112:125. 

Morse, R.A., D.M. Burgett, J.T. Ambrose, W.E. Conner and R.D. Fell. 

1973. Early introductions of African bees into Europe and the New 
World. Bee World 54:57-60. 

Nelson, J. A., and A. P. Sturtevant. 1924. The rate of growth of the 

honeybee larvae. Department Bulletin 1222. USDA, Washington, D.C 

Nolan, W.J. 1925. The brood-rearing cycle of the honeybee. Department 
Bulletin 1349. USDA, Washinton, D.C. 

Nolan, W.J. 1928. Seasonal brood-rearing activity of the Cyprian 
honeybee. J. Econ. Ent. 21:392-403. 

Nunamaker, R.A., and W.T. Wilson. 1981. Comparison of MDH allozyme 

patterns in the African honey bee ( Apis mellifera adansonii L.) and 
the Africanized populations of Brazil. J. Kans. Ent. Soc. 54: 
704-710. 

Nunez, J. A. 1973. Quantitative investigation of the behaviour of Apis 
mglHf^ra liqiustica Spinola and Apis mellifera adansonii Latreille: 
Energy factors, forager recruitment and foraging activity. Apiacta 
8:151-154. 



175 

Nunez» J. A, 1979a. Times spent on various components of foraging 

activity: Comparison between European and Africanized honeybees in 
Brazil. J. Apic. Res. 18:110-115. 

Nunez* J. A. 1979b. Comparative study of thermoregulation between 

European and Africanized Apis mell if era in Brazil. J. Apic. Res. 
18:116-121. 

Nunez* J. A. 1982. Comparacion del comportamiento recolector de abejas 

African izadas y abejas Europeas. Pages 221-231 in P. Jaisson 

(ed.), Social insects in the tropics. Universite Paris-Nord» 
Paris. 

Oertel» E. 1940. Mating flights of queen bees. Glean. Bee Cult. 
68:292-293. 

Otis, G.W. 1980. The swarming biology and population dynamics of the 
Africanized honeybee. Ph.D. Dissertation, Univ. of Kansas, 
Lawrence. 

Otis, G.W. 1982a. Population biology of the Africanized honey bee. 
Pages 209-219 in P. Jaisson (ed.). Social insects in the tropics. 
Universite Paris-Nord, Paris. 

Otis, G.W. 1982b. Weights of worker honeybees in swarms. J. Apic. 
Res. 21:88-92. 

Otis, G.W., and O.R. Taylor. 1979. Beekeeping in the Guianas. Pages 
145-154 in Beekeeping in rural development: Unexploited beekeeping 
potential in the tropics. Commonwealth Secretariat, London. 

Otis, G.W., M.L. Winston and O.R. Taylor. 1981. Engorgement and 

dispersal of Africanized honeybee swarms. J. Apic. Res. 20:3-12. 

Page, R.E., Jr., and E.H. Erickson, Jr. 1984. Selective rearing of 
queens by worker honey bees: Kin or nestmate recognition. Ann. 
Ent. Soc. Am. 77:578-580. 

Page, R.E., Jr., and E.H. Erickson, Jr. 1985. Identification and 
certification of Africanized honey bees. Ann. Ent. Soc. Am. 
78:149-158. 

Page, R.E., Jr., and R.A. Metcalf. 1982. Multiple mating, sperm 
utilization and social evolution. Am. Nat. 119:263-281. 

Peer, D.F. 1956. Multiple mating of queen honey bees. J. Econ. Ent. 
49:741-743. 

Pellett, F.C. 1938. History of American beekeeping. Collegiate Press, 
Inc., Ames, Iowa. 

Reid, M. 1975. Storage of queen honeybees. Bee World 56:21-31. 



176 

Ribbands, C.R. 1953. The behaviour and soda! Ufa of honeybees. Bee 
Research Assoc, Ltd., England. 

Rinderer, T.E. 1986. Afrlcam'zed bees: An overview. Am. Bee J. 
126:98-100, 128-129. 

Rinderer, T.E., A.B. Bolten.^A.M. Collins and J.R. Harbo. 1984. 
Nectar-foraging characteristics of Africanized and European 
honeybees in the neotropics. J. Apic. Res. 23:70-79. 

Rinderer, T.E., A.B. Bolten, J.R. Harbo and A.M. Collins. 1982. 
Hoarding behavior of European and Africanized honey bees 
(Hymenoptera: Apidae). J. Econ. Ent. 75:714-715. 

Rinderer, T.E., A.M. Collins, A.B. Bolten and J.R. Harbo. 1981. Size 
of nest cavities selected by swarms of Africanized honeybees in 
Venezuela. J. Apic. Res. 20:160-164. 

Rinderer, T.E., A.M. Collins and K.W. Tucker. 1985. Honey production 
and underlying nectar harvesting activities of Africanized and 
European honeybees. J. Apic. Res. 24:161-167. 

Rinderer, T.E., R.L. Hellmich, II, R.G. Danka and A.M Collins. 1985. 
Male reproductive parasitism: A factor in the Africanization of 
European honey-bee populations. Science 228:1119-1121. 

Rinderer, T.E., and H.A. Sylvester. 1981. Identification of 
Africanized bees. Am. Bee J. 121:512-516. 

Rinderer, T.E., K.W. Tucker and A.M. Collins. 1982. Nest cavity 
selection by swarms of European and Africanized honeybees. J. 
Apic. Res. 21:98-103. 

Roberts, W.C. 1944. Multiple mating of queen bees proved by progeny 
and flight tests. Glean. Bee Cult. 72:255-259, 303. 

Roberts, W.C, and S. Taber. 1965. Egg-weight variance in honey bees. 
Ann. Ent. Soc. Am. 58:303-306. 

Root, A.I. 1947. ABC and XYZ of bee culture. Root Co., Medina, Ohio. 

Roublk, D.W. 1978. Competitive interactions between neotropical 

pollinators and Africanized honey bees. Science 201:1030-1032. 

Roublk, D.W. 1979. Africanized honeybees, stingless bees and the 

structure of tropical plant-pollinator communities. Pages 403-417 
in D. Caron (ed.), Proc. IVth Intl. Symp. on Pollination. Misc. 
Publ. no. 1, Maryland Agrlc. Exp. Sta., Univ. of Maryland, College 
Park. 

Roublk, D.W. 1980. Foraging behavior of competing Africanized 
honeybees and stingless bees. Ecology 61:836-845. 



177 

Roublk, D.W. 1982. Ecological Impact of Africanized honeybees on 

native neotropical pollinators. Pages 233-247 in P. Jalsson (ed,)» 
Soda! Insects In the tropics. Universlte Par1s-Nord» Paris. 

Roub1k» D.W. 1983. Experimental community studies: Time-series tests 
of competition between African and neotropical bees. Ecology 
64:971-978. 

Roub1k» D.W.» and S.L. Buchmann. 1984. Nectar selection by Mell Ipona 
and A p1s mell If era (Hymenoptera: Apldae) and the ecology of nectar 
Intake by bee colonies In a tropical forest. Oecologia 61:1-10. 

Ruttner* F. 1968. Les races d'abellles. Pages 27-44 In R. Chauvin 

(ed.)» Tralte de blologle de l'abenies» 1. Masson et C1e» Paris. 

Ruttner, F. 1975. Races of bees. Pages 19-38 in Dadant and Sons 
(eds.)» The hive and the honey bee. Dadant and Sons» Inc.* 
Hamilton, Illinois. 

Ruttner, F. 1976a. The races of bees of Africa. Proc. 25th Int. 

Beekeep. Congr., 1975, Grenoble. Aplmondia, Bucharest, Rumania. 

Ruttner, F. 1976b. Honeybees of the tropics: Their variety and 

characteristics of Importance for apiculture. Pages 41-46 In E. 
Crane (ed,). Apiculture In tropical climates. International Bee 
Research Assoc, London. 

Ruttner, F. 1981. On the taxonomy of honey bees of tropical Africa. 
Pages 278-287 in Proc. 28th Int. Congr. Apimondia, Acapulco, 
Mexico. 

Seeley, T.D. 1977. Measurement of nest cavity volume by the honey bee 
( Apis mell if era ). Behav. Ecol. Sociobiol. 2:201-227. 

Seeley, T.D. 1978. Life history strategy of the honey bee. Apis 
mell if era . Oecologia 32:109-118. 

Seeley, T.D. 1979. Queen substance dispersal by messenger workers in 
honeybee colonies. Behav. Ecol. Sociobiol. 5:391-415. 

Seeley, T.D. 1982. Adaptive significance of the age polyethism 

schedule in honeybee colonies. Behav. Ecol. Sociobiol. 11:287-293. 

Seeley, T.D. 1983. The ecology of temperate and tropical honeybee 
societies. Am. Sci. 71:264-272. 

Seeley, T.D., and R.D. Fell. 1981. Queen substance production in honey 
bee ( Apis mell If era ) colonies preparing to swarm (Hymenoptera: 
Apldae). J. Kans. Ent. Soc. 54:192-196. 

Seeley, T.D., and R.A. Morse. 1976. The nest of the honey bee ( Apis 
mell if era L.). Insectes Sociaux 23:495-512. 



178 

Seeley, T.D., R.H. Seeley and P. Akratanakul. 1982. Colony defense 
strategies of the honeybees in Thailand. Ecol. Monogr. 52:43-63. 

Shearer* D.A., R. Boch» R.A. Morse and F.M. Laigo. 1970. Occurrence of 
9-oxodec-trans-2-enoic acid in queens of A pjs dorsata » Apis cerana 
and Apis mell if era . J. Insect Physiol. 16:1437-1441. 

Siegel, S. 1956. Nonparametric statistics for the behavioral sciences. 
McGraw-Hill Book Co., New York. 

Simpson, J. 1966. Congestion of adult honeybees with and without 
adequate hive space. J. Apic. Res. 5:59-61. 

Simpson, J. 1973. Influence of hive-space restriction on the tendency 
of honeybee colonies to rear queens. J. Apic. Res. 12:183-186. 

Simpson, J., and I.B.M. Riedel. 1963. The factor that causes swarming 
by honeybee colonies in small hives. J. Apic. Res. 2:50-54. 

Smith, F.G. 1958a, Beekeeping observations in Tanganyika 1949-1957. 
Bee World 39:29-36. 

Smith, F.G. 1958b. Communication and foraging range of African bees 
compared with that of European and Asian bees. Bee World 39: 
249-252. 

Smith, F.G. 1961. The races of honeybees in Africa. Bee World 42: 
255-260. 

Smith, M.V. 1972. Marking bees and queens. Bee World 53:9-13. 

Sokol, R.R., and F.J. Rohlf. 1969. Biometry. W.H. Freeman and Co., 
San Francisco. 

Stort, A.C. 1974. Genetic study of aggressiveness of two subspecies of 
Apis mell if era in Brazil, 1. Some tests to measure aggressiveness. 
J. Apic. Res. 13:33-38. 

Stort, A.C. 1975a, Genetic study of aggressiveness of two subspecies 
of Apis mell if era in Brazil, 2. Time at which the first sting 
reached a leather ball. J. Apic. Res. 14:171-175. 

Stort, A.C. 1975b. Genetic study of the aggressiveness of two 

subspecies of Apis mell if era in Brazil, IV. Number of stings in the 
gloves of the observer. Behavior Genetics 5:269-274. 

Stort, A.C. 1975c. Genetical study of the aggressiveness of two 

subspecies of Apis mellifera in Brasil, V. Number of stings in the 
leather ball. J. Kans. Ent. Soc. 48:381-387. 

Stort, A.C. 1976. Genetic study of the aggressiveness of two 

subspecies of Apis mell if era in Brazil, III. Time taken for the 
colony to become aggressive. Ciencia e Cultura 28:1182-1185. 



179 

Stort» A.C., and M, BarelH. 1981. Genetic study of olfactory 

structures 1n the antennae of two Apis mell -if era subspecies. J. 
Kans. Ent. Soc. 54:352-358. 

Sylvester» H.A. 1982, Electrophoretic identification of Africanized 
honeybees. J. Apic. Res. 21:93-97. 

Taber* S.» III. 1954. The frequency of multiple mating of queen honey 
bees. J. Econ. Ent. 47:995-998. 

Taber. S., III. 1961. Successful shipments of honeybee semen. Bee 
World 42:173-176. 

Taber» S., Ill, and W.C. Roberts. 1963. Egg weight variability and its 
inheritance in the honey bee. Ann. Ent. Soc. Am. 56:473-476. 

Taber» S., Ill, and J. Wendel. 1958. Concerning the number of times 
queen bees mate. J. Econ. Ent. 51:786-789. 

Taylor* O.R. 1977. The past and possible future spread of Africanized 
honeybees in the Americas. Bee World 58:19-30. 

Taylor, O.R. 1985. African bees: Potential impact in the United 
States. Bull. Ent. Soc. Am. 31:14-24. 

Taylor* O.R. 1986. Health problems associated with African bees. Ann. 
Int. Med. 104:267-268. 

Taylor, O.R., R.W. Kingsolver and G.W. Otis. In press. A neutral 
mating model for honey bees ( Apis mell if era L.). J. Apic. Res. 

Taylor, O.R., and M. Spivak. 1984. Climatic limits of tropical African 
honeybees in the Americas. Bee World 65:38-47. 

Taylor, O.R., and G.B. Williamson. 1975. Current status of the 
Africanized honey bee in northern South America. Am. Bee J. 
115:92-93, 98-99. 

Tribe, 6.D., and D.J.C. Fletcher. 1977. Rate of development of the 
workers of Apis mell if era adansonii L. Pages 115-119 in D.J.C. 
Fletcher (ed.), African bees: Taxonomy, biology and economic use. 
Apimondia, Pretoria. 

Tuenin, T.A. 1927. Variation of bees and their organs. Am. Bee J. 
67:19. 

Visscher, P.K., and T.D. Seeley. 1982, Foraging strategy of honeybee 
colonies in a temperate deciduous forest. Ecology 63:1790-1801. 

von Frisch, K. 1967. The dance language and orientation of bees. 
Belknap Press, Cambridge, Mass. 

Waddington, K.D. 1981. Patterns of size variation in bees and 
evolution of communication systems. Evolution 35:813-814. 



180 

Waddington, K.D., L.H. Herbst and D.W. Roublk. 1986. Relationship 
between recruitment systems of sting! ess bees and within-nest 
worker size variation. J. Kans. Ent. Soc. 59:95-102. 

Wafa» A.K., S.E. Rashad and M.M. Mazeed. 1965. Biometrical studies on 
the Egyptian honeybee. J. Ap1c. Res. 4:161-166. 

Weiss, K. 1974. Zur frage des koniginnengewichtes in abhangigkeit von 
umlarvalter und larvenversorgung. Apidologie 5:127-147. 

Weiss, K, 1984. ReguTierung des proteinhaushaltes im bienenvolk ( Apis 
men if 1ca L.) durch brutkannibal ismus. Apidologie 15:339-354. 

Wenner, A.M. 1962. Sound production during the waggle dance of the 
honey bee. Anim. Behav. 10:79-95. 

Wiese, H. 1972. Abel has africanas, suas caracteristicas e tecnologia 

de manejo. Pages 95-108 in Anais do 1° Congresso Brasileiro de 
Apicultura. Congresso Brasileiro de Apicultura, Florianopol is, 
Brazil. 

Wilson, E.G. 1971. The Insect societies. Belknap Press, Cambridge, 
Mass. 

Winston, M.L. 1979a. The potential Impact of the Africanized honey bee 
on apiculture in Mexico and Central America. Am. Bee J. 119:584- 
586, 642-645. 

Winston, M.L. 1979b. Intra-colony demography and reproductive rate of 
the Africanized honeybee in South America. Behav. Ecol. Sociobiol. 
4:279-292. 

Winston, M.L. 1979c. Events following queen removal in colonies of 
Africanized honey bees in South America. Insectes Soclaux 26: 
373-381. 

Winston, M.L. 1980a. Swarming, afterswarmlng and reproductive rate of 
unmanaged honeybee colonies ( Apis mellifera ). Insectes Sociaux 
27:391-398. 

Winston, M.L. 1980b. Seasonal patterns of brood rearing and worker 
longevity in colonies of the Africanized honey bee (Hymenoptera: 
Apidae) in South America. J. Kans. Ent. Soc. 53:157-165. 

Winston, M.L., J. Dropkin and G.R. Taylor. 1981. Demography and life 
history characteristics of two honey bee races ( Apis mellifera ). 
Oecologia 48:407-413. 

Winston, M.L., and S.J. Katz. 1981. Longevity of cross-fostered honey 
bee workers ( Apjg mellifera ) of European and Africanized races. 
Can. J. Zool. 59:1571-1575. 



181 

Winston* M.L.* and S.J. Katz. 1982. Foraging differences between 

cross-fostered honeybee workers ( Apis mell if era ) of European and 
Africanized races. Behav. Ecol . Sociobiol. 10:125-129. 

Winston, M.L.» and G.W. Otis. 1978. Ages of bees in swarms and 

afterswarms of the Africanized honeybee. J. Apic. Res. 17:123-129. 

Winston, M.L., G.W. Otis and O.R. Taylor. 1979. Absconding behaviour 
of the Africanized honeybee in South America. J. Apic. Res. 18:85- 
94. 

Winston, M.L., and O.R. Taylor. 1980. Factors preceding queen rearing 
in the Africanized honeybee ( Apis mell if era ) in South America. 
Insectes Sociaux 27:289-304. 

Winston, M.L., O.R. Taylor and G.W. Otis. 1983. Some differences 

between temperate European and tropical African and South American 
honeybees. Bee World 64:12-21. 

Woyke, J. 1969. African honey bees in Brazil. Am. Bee J. 109:342-344. 

Woyke, J. 1971. Correlations between the age at which honeybee brood 
was grafted, characteristics of the resultant queens, and results 
of insemination. J. Apic. Res. 10:45-55. 

Woyke, J. 1973. Experiences with Apis mell if era adansonii in Brazil 
and in Poland. Apiacta 8:115-116. 

Woyke, J. 1976. Brood-rearing efficiency and absconding in Indian 
honeybees. J. Apic. Res. 15:133-143. 

Woyke, J. 1977. Cannibalism and brood-rearing efficiency in the 
honeybee. J. Apic. Res. 16:84-94. 

Woyke, J. 1979. Effects of the access of worker honeybees to the queen 
on the results of the instrumental insemination. J. Apic. Res. 
19:136-143. 

Woyke, J. 1984. Correlations and interactions between population, 
length of worker life and honey production by honeybees in a 
temperate region. J. Apic. Res. 23:148-156. 

Zeuner, F.E., and F.J. Manning. 1976. A monograph on fossil bees 
(Hymenoptera: Apoidea). Bull. Brit. Mus. (Nat. Hist.), Geol. 
27:151-268. 

Zmarlicki, C, and R.A. Morse. 1963. Drone congregation areas. J. 
Apic, Res. 2:64-66. 



BIOGRAPHICAL SKETCH 

Alan Bolten was born on May 27, 1945 In Newark, New Jersey. In 
1959, he was graduated from Maple Avenue Grammar School. Four years 
later, he completed his secondary education at Weequahic High School in 
Newark. He received his undergraduate education from Union College in 
Schenectady, New York, where he was graduated with honors in biology in 
1967. Alan began graduate studies in the Department of Zoology at the 
University of Florida in 1977. He is married to Karen Bjorndal. 



182 



I certify that I have read this study and that in my opinion it 
conforms to acceptable standards of scholarly presentation and is fully 
adequate, in scope and quality, as a dissertation for the degree of 
Doctor of Philosophy. 



£>-^ 





Thomas C. Emmel, Chairman 
Professor of Zoology 



I certify that I have read this study and that in my opinion it 
conforms to acceptable standards of scholarly presentation and is fully 
adequate, in scope and quality, as a dissertation for the degree of 
Doctor of Philosophy. 




mtlmn Reiskind 
Associate Professor of Zoology 



I certify that I have read this study and that in my opinion it 
conforms to acceptable standards of scholarly presentation and is fully 
adequate, in scope and quality, as a dissertation for the degree of 
Doctor of Philosophy. / /~' 

Malcolm T. Sanf^iard ~' ' 

Associate Prof essor £i'f Entomology 
and Nematology 



This dissertation was submitted to the Graduate Faculty of the 
Department of Zoology in the College of Liberal Arts and Sciences and to 
the Graduate School and was accepted as partial fulfillment of the 
requirements for the degree of Doctor of Philosophy, 



August 1986 



Dean, Graduate School 



n i 



A 



/^